diff --git a/.gitattributes b/.gitattributes index bed0738c7eeb449bca98b5d2f33c89a1ee56349a..7fb602ff01259bcd78b753a8561e626fda381e6d 100644 --- a/.gitattributes +++ b/.gitattributes @@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text # Video files - compressed *.mp4 filter=lfs diff=lfs merge=lfs -text *.webm filter=lfs diff=lfs merge=lfs -text +results/omni_finetune/adapter_lora/tokenizer.json filter=lfs diff=lfs merge=lfs -text +results/omni_finetune/hf_upload/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/ARTIFACT_GUIDE.md b/ARTIFACT_GUIDE.md index f9ca2bc7fe112aed8efeb3f8fe3f28e22eea5aa6..2f90ebab050fd7da67c25d5daaa03b51b3b17b8e 100644 --- a/ARTIFACT_GUIDE.md +++ b/ARTIFACT_GUIDE.md @@ -8,8 +8,8 @@ The project intentionally separates nine layers: 1. **Reviewer scorecard:** one compact table for first-pass current-state decisions. -2. **Proof boundary:** what is claimed, what is smoke-only, and what remains - gated by data access. +2. **Proof boundary:** what is claimed, what is readiness-only, and what + remains gated by data access. 3. **Official source alignment:** what the upstream Xperience-10M dataset card, public sample card, and HF API metadata say, and which parts this repo currently covers. @@ -128,7 +128,7 @@ The project intentionally separates nine layers: | Artifact | Current status | | --- | --- | | [`results/omni_finetune/DATA_BLOCKER_REPORT.md`](results/omni_finetune/DATA_BLOCKER_REPORT.md) | Documents why no real 32-episode Qwen3-Omni result is claimed yet. | -| [`results/omni_finetune/A100_HF_RELAY_STATUS.md`](results/omni_finetune/A100_HF_RELAY_STATUS.md) | Documents the pending A100-to-H20 relay and selected 32-session pilot plan. | +| [`results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md`](results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md) | Documents the public multi-episode access boundary and selected 32-episode pilot plan without private infrastructure details. | | [`scripts/omni/discover_xperience10m_sources.py`](scripts/omni/discover_xperience10m_sources.py) | Discovery gate for valid multi-episode Xperience-10M sources. | | [`scripts/omni/train_qwen3_omni_lora.py`](scripts/omni/train_qwen3_omni_lora.py) | Training entrypoint for the Qwen3-Omni LoRA pilot after the data gate passes. | diff --git a/EVALUATION_PROTOCOL.md b/EVALUATION_PROTOCOL.md index 4e6318141be1db98bacde01945e91f3c60f8b7fb..c7a25bed459ce025db43b121662798fded5fb545 100644 --- a/EVALUATION_PROTOCOL.md +++ b/EVALUATION_PROTOCOL.md @@ -72,7 +72,7 @@ are not foundation models. - Do not infer cross-episode generalization from this single public sample. - Do not treat feature-vector reconstruction as pixel depth, mesh, NeRF, or Gaussian reconstruction. -- Do not treat Qwen3-Omni smoke artifacts as a real 32-episode fine-tune. +- Do not treat Qwen3-Omni readiness artifacts as a real 32-episode fine-tune. - Do not infer audio-visual learning from the current baseline vector because audio is not featurized. ## Scale-Up Gate @@ -87,7 +87,7 @@ being presented as model quality: Current status: prepared but data-gated. Read `results/omni_finetune/DATA_BLOCKER_REPORT.md` and -`results/omni_finetune/A100_HF_RELAY_STATUS.md` before interpreting any +`results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md` before interpreting any Qwen3-Omni artifact. ## Machine-Readable Copy diff --git a/EVIDENCE_CONTRACT.md b/EVIDENCE_CONTRACT.md index 3b92b00079cbfbe3ac99e9be43522e4d78dbb951..333c89b60c1d90ba5322fcdb7fe8ed05d83c850f 100644 --- a/EVIDENCE_CONTRACT.md +++ b/EVIDENCE_CONTRACT.md @@ -18,8 +18,8 @@ local artifact that a reader can inspect before trusting the dashboard. | Minimal and neural heads use the same task contracts. | `scripts/neural_task_models.py`, `results/episode_task_suite/neural_mlp/`, `docs/assets/task_architectures.png` | Verified for 12 minimal heads and 12 neural MLP heads | Small heads only; not a foundation model | | Four Ropedia research directions are mapped honestly as direct, proxy, or diagnostic evidence. | `results/episode_task_suite/research_directions/research_direction_taxonomy.json`, `docs/data/research_directions.json` | Verified taxonomy | Some directions remain proxy-only | | Four extra direction probes are coded and evaluated. | `results/episode_task_suite/research_direction_extensions/research_direction_extension_results.json`, `docs/data/research_direction_extensions.json` | Verified single-episode probes | Not full human modeling, neural rendering, intent modeling, or world modeling solutions | -| Qwen3-Omni infrastructure has passed technical smoke checks. | `results/omni_finetune/RUN_REPORT.md`, `results/omni_finetune/dataset_manifest.json`, `results/omni_finetune/metrics_eval.json` | Smoke-only evidence | One episode, 128 train windows; not a 32-episode pilot | -| The real 32-episode LoRA pilot is blocked on gated data access, not on repo presentation. | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/A100_HF_RELAY_STATUS.md`, `results/omni_finetune/source_discovery.json` | Blocker documented | No 32-episode metric should be claimed until the gate passes | +| Qwen3-Omni infrastructure has passed readiness checks. | `results/omni_finetune/RUN_REPORT.md`, `results/omni_finetune/dataset_manifest.json`, `results/omni_finetune/metrics_eval.json` | Readiness-only evidence | One episode, 128 train windows; not a 32-episode pilot | +| The real 32-episode LoRA pilot is blocked on gated data access, not on repo presentation. | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md`, `results/omni_finetune/source_discovery.json` | Blocker documented | No 32-episode metric should be claimed until the gate passes | | Historical `32ep` path strings are not treated as 32-episode results. | `scripts/validate_scope_claims.py`, `docs/data/scope_claims_audit.json` | Verified pass | Classifies old run/path identifiers and fails if public presentation claims real 32-episode metrics | | Prepared GitHub/Hugging Face mirrors carry matching critical files. | `scripts/validate_mirror_parity.py`, `docs/data/mirror_parity.json` | Verified pass | Compares prepared data files, visual assets, website HTML, and validator scripts before upload; live URLs are checked after publishing | | The public GitHub and Hugging Face bundles are publication-clean. | `scripts/validate_publication_package.py`, `docs/data/publication_audit.json` | Verified pass | Checks public files, HF bundles, and public-card freshness; ignored local scratch outputs are excluded | @@ -63,11 +63,12 @@ local artifact that a reader can inspect before trusting the dashboard. 12. Inspect `results/episode_task_suite/neural_mlp/` to compare minimal and neural heads under the same splits. 13. Inspect `docs/data/scope_claims_audit.json` before interpreting historical - `32ep` strings in Qwen3-Omni smoke artifacts. + `32ep` strings in Qwen3-Omni readiness artifacts. 14. Inspect `docs/data/mirror_parity.json` before assuming the GitHub and Hugging Face mirrors contain the same critical data, visual, HTML, and validator files. -15. Inspect `results/omni_finetune/DATA_BLOCKER_REPORT.md` before interpreting +15. Inspect `results/omni_finetune/DATA_BLOCKER_REPORT.md` and + `results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md` before interpreting any Qwen3-Omni artifact. 16. Inspect `QUALITY_GATES.md`, `docs/data/quality_gates.json`, `docs/data/publication_audit.json`, and `docs/data/website_integrity.json` diff --git a/PROJECT_README.md b/PROJECT_README.md index c0a9358a281c986d8bb80d89eac0a34e63fc413b..2ddcd0de3208f770acd305ddf17956b5017d0930 100644 --- a/PROJECT_README.md +++ b/PROJECT_README.md @@ -14,12 +14,6 @@ An audit-first embodied-AI learning repo built around one public Xperience-10M sample episode released by Ropedia. -The public dashboard and generated figures deliberately follow the visual -language of [ropedia.com](https://ropedia.com/): near-black 4D-world canvas, -lime-green identity accents, thin green-tinted cards, point-cloud texture, and -the Inter Tight / Space Grotesk typography pairing. The layout is original to -this project, but the style stays aligned with Ropedia's own product site. - The project does one narrow thing carefully: it turns a raw multimodal episode into: @@ -50,7 +44,7 @@ This repo is organized around an explicit proof boundary: | 12-task suite | `scripts/episode_task_suite.py`, per-task `metrics.json`, predictions | chronological single-episode split | | Neural heads | `scripts/neural_task_models.py`, `results/episode_task_suite/neural_mlp/` | compact MLP heads, not a foundation model | | Research directions | `research_direction_taxonomy.json`, extension probe results | direct/proxy/diagnostic evidence, not full solutions | -| Qwen3-Omni | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `A100_HF_RELAY_STATUS.md` | smoke-only until 32 valid episodes are available | +| Qwen3-Omni | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `MULTI_EPISODE_ACCESS_STATUS.md` | readiness-only until 32 valid episodes are available | | Scope claims guard | `scripts/validate_scope_claims.py`, `docs/data/scope_claims_audit.json` | historical `32ep` path strings are provenance, not 32-episode results | | Mirror parity | `scripts/validate_mirror_parity.py`, `docs/data/mirror_parity.json` | prepared GitHub/HF mirrors carry matching data, figure, website HTML, and validator files | | Publication hygiene | `scripts/validate_publication_package.py`, `docs/data/publication_audit.json` | public repo and HF bundles only; ignored local scratch files are excluded, and public cards must reference the current task-first figure | @@ -135,7 +129,7 @@ If you are reviewing the project cold, open these in order: | 6 | What is one model input? | [`windows.csv`](results/episode_task_suite/windows.csv), [`feature_manifest.json`](results/episode_task_suite/feature_manifest.json), [`available_modalities.json`](results/episode_task_suite/available_modalities.json) | The input is an aligned 8,378-d window vector with explicit feature-block boundaries. | | 7 | Are the task results backed by files? | [`summary_report.json`](results/episode_task_suite/summary_report.json), [`neural_mlp/`](results/episode_task_suite/neural_mlp/), [`docs/data/summary_metrics.json`](docs/data/summary_metrics.json) | Each task has minimal and neural-head evidence over the same window contracts. | | 8 | Is the website internally coherent? | [`docs/data/website_integrity.json`](docs/data/website_integrity.json), [`scripts/validate_website_integrity.py`](scripts/validate_website_integrity.py) | Local links, anchors, JSON data, and referenced images are checked before publishing. | -| 9 | What is still pending? | [`DATA_BLOCKER_REPORT.md`](results/omni_finetune/DATA_BLOCKER_REPORT.md), [`A100_HF_RELAY_STATUS.md`](results/omni_finetune/A100_HF_RELAY_STATUS.md), [`scripts/omni/discover_xperience10m_sources.py`](scripts/omni/discover_xperience10m_sources.py) | The 32-episode Qwen3-Omni run is prepared but not yet a real model-quality claim. | +| 9 | What is still pending? | [`DATA_BLOCKER_REPORT.md`](results/omni_finetune/DATA_BLOCKER_REPORT.md), [`MULTI_EPISODE_ACCESS_STATUS.md`](results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md), [`scripts/omni/discover_xperience10m_sources.py`](scripts/omni/discover_xperience10m_sources.py) | The 32-episode Qwen3-Omni run is prepared but not yet a real model-quality claim. | The machine-readable reviewer packet is [`docs/data/reviewer_packet.json`](docs/data/reviewer_packet.json). @@ -333,13 +327,13 @@ scripts/ build_artifact_index.py # builds the source-of-truth reviewer index build_quality_gates.py # builds reviewer-facing publication gates validate_mirror_parity.py # checks prepared GitHub/HF mirror file parity - validate_scope_claims.py # checks Qwen3-Omni smoke/result claim boundaries + validate_scope_claims.py # checks Qwen3-Omni readiness/result claim boundaries validate_website_integrity.py # checks local site links, anchors, JSON, images validate_publication_package.py # checks public repo + HF bundle hygiene omni/ - download_sample_modelscope.py # mainland-China friendly sample download + download_sample_modelscope.py # ModelScope sample download helper build_episode_manifest.py # metadata-only multi-episode scanner - plan_finetune_sample_budget.py # H20 storage/sample-count planner + plan_finetune_sample_budget.py # storage/sample-count planner qwen3_omni_adapter_smoke.py # real-data Qwen3-Omni adapter smoke test results/ @@ -352,7 +346,7 @@ results/ research_directions/ # four-track taxonomy, CSV, and summary research_direction_extensions/ # four extra direction probes + predictions task_walkthroughs/ # case-study walkthroughs for all 12 tasks - omni_exploration/ # H20/ModelScope smoke-test artifacts + omni_exploration/ # ModelScope readiness-check artifacts docs/ index.html # GitHub Pages dashboard @@ -434,7 +428,7 @@ hf download ropedia-ai/xperience-10m-sample \ --local-dir data/sample/xperience-10m-sample ``` -On mainland-China servers, use ModelScope instead: +If Hugging Face access is unavailable in your environment, use ModelScope: ```bash python scripts/omni/download_sample_modelscope.py \ @@ -470,111 +464,52 @@ python scripts/train_min_action_model.py --workspace /path/to/workspace python scripts/train_all_modalities_model.py --workspace /path/to/workspace ``` -## Xperience-10M Fine-Tuning Exploration On H20 +## Xperience-10M Fine-Tuning Exploration -This repo now includes a concrete first step toward a Qwen3-Omni fine-tuning -pipeline over Xperience-10M. The important separation is: +This repo includes a first Qwen3-Omni fine-tuning path over Xperience-10M, but +the current evidence is still readiness evidence rather than model quality. +The useful distinction is: - direct Qwen3-Omni inputs: RGB/fisheye video, embedded MP4 audio, and language prompts, - adapter-required Xperience-10M sensor inputs: depth, pose/SLAM, hand/body mocap, contacts, and IMU. -The H20 work now has two separate evidence levels: - -- an adapter-side smoke test over one Xperience-10M sample episode, useful for - checking sensor feature extraction and label plumbing, -- a technical Qwen3-Omni LoRA smoke run that loaded the local - `Qwen/Qwen3-Omni-30B-A3B-Instruct` weights and trained LoRA parameters on - 128 windows from the single locally available episode. - -Neither is a 32-episode result. The full pilot is still gated on raw -Xperience-10M access and a held-out episode split. - -```bash -python scripts/omni/build_episode_manifest.py \ - --data-root /home/cy/Ropedia/modelscope_data \ - --output outputs/omni_exploration/modelscope_manifest.json - -python scripts/omni/qwen3_omni_adapter_smoke.py \ - --workspace /home/cy/Ropedia/ropedia-xperience-10m-task-suite \ - --episode-root /home/cy/Ropedia/modelscope_data/xperience-10m-sample \ - --target action \ - --window-frames 20 \ - --stride-frames 100 \ - --max-windows-per-episode 64 \ - --epochs 2 \ - --skip-video-features -``` - -Verified H20 run: - -| Item | Value | -| --- | ---: | -| Server | 8 x NVIDIA H20, 96GB each | -| Free storage checked | about 1.5TB under `/home/cy` | -| Data source | ModelScope `ropedia-ai/xperience-10m-sample` | -| Downloaded minimal data | 1.93GB `annotation.hdf5` + 85.7MB `fisheye_cam0.mp4` | -| Smoke windows | 59 | -| Split | single-episode chronological | -| Feature dim | 4,262 | -| Adapter soft-token blocks | 11 | -| Qwen3-Omni weights loaded | adapter smoke: no; LoRA smoke: yes | -| Result | 0.0000 macro-F1, expected for this single-episode chronological smoke split | - -The zero score is not treated as a model claim. It is a useful signal that this -split is not leaking labels across time: the train segment does not cover every -action that appears in the held-out segment. The next real step is to add more -episodes and split by held-out episode. +The current scale-up artifacts prove that the export, manifest, sensor-feature, +LoRA, and evaluation scripts can run on the available sample episode. They do +not prove a real 32-episode result. A real pilot requires at least 32 valid +episodes, held-out episode splits, training metadata, predictions, metrics, and +a run report. ### Sample Count Decision -The local Mac sample is only one episode. For H20 fine-tuning, decide sample -count by storage and evaluation design, not by the local folder. The current H20 -has about 1.5TB free under `/home/cy`; after reserving space for model weights, -checkpoints, caches, and logs, a realistic first budget is: +Do not treat "10M" as a reason to start with the entire dataset. The engineering +unit that matters first is diverse held-out episodes, not adjacent windows from +one session. | Phase | Episodes/samples | Approx windows at stride 5 | Purpose | | --- | ---: | ---: | --- | -| Smoke | 1-3 | 1k-3k | Verify loaders, token alignment, and task heads | +| Readiness | 1-3 | 1k-3k | Verify loaders, token alignment, and task heads | | Pilot | 16-32 | 18k-37k | First held-out-episode evaluation | | Useful LoRA run | 64-128 | 74k-149k | Train sensor adapters plus selected Qwen3-Omni LoRA | | Storage-heavy run | 256+ | 297k+ | Only after download layout and checkpoint size are stable | -For the next run, use a **32-episode stratified pilot** through the A100 relay, -then scale to **128 episodes** and later **512 episodes** only after the -download, transfer, manifest, train, and held-out evaluation path is stable. Do -not treat "10M" as a reason to start with the entire dataset; the engineering -unit that matters first is diverse held-out episodes, not adjacent windows from -one session. - Use the budget helper before downloading: ```bash python scripts/omni/plan_finetune_sample_budget.py \ - --storage-root /home/cy \ + --storage-root /path/to/storage \ --target-free-after-download-gb 800 \ --all-training-per-episode-gb 2.4 \ --full-preview-per-episode-gb 5.1 ``` -Refresh charts and the website data bundle: - -```bash -python scripts/research_direction_taxonomy.py -python scripts/research_direction_extension_tasks.py -python scripts/task_walkthroughs.py -python scripts/generate_visualizations.py -python scripts/render_overview_figures.py -python scripts/render_task_suite_infographic.py -``` - ### 32-Episode Readiness Gate ```bash python scripts/omni/discover_xperience10m_sources.py \ - --workspace /home/cy/Ropedia/ropedia-xperience-10m-task-suite \ - --data-root /home/cy/Ropedia/modelscope_data \ + --workspace /path/to/ropedia-xperience-10m-task-suite \ + --data-root /path/to/xperience10m_data \ --output results/omni_finetune/source_discovery.json \ --report-output results/omni_finetune/DATA_BLOCKER_REPORT.md ``` @@ -584,33 +519,17 @@ Current status in this repo: - local_valid_episodes: 1 (degraded-valid: annotation + fisheye_cam0.mp4) - local_complete_episodes: 0 - ready_for_32_episode_pilot: false -- A100 Hugging Face relay: active watcher, polling gated access every 15 minutes - planned 32-episode pilot: stratified across 32 top-level session UUIDs -- HF full dataset blocker: `ropedia-ai/xperience-10m` returns 403 pending review +- full-dataset blocker: gated Xperience-10M access is still pending - source_discovery: `results/omni_finetune/source_discovery.json` - blocker_report: `results/omni_finetune/DATA_BLOCKER_REPORT.md` -- relay_status: `results/omni_finetune/A100_HF_RELAY_STATUS.md` - -Current H20-sourced evidence files in this repo: - -- `results/omni_finetune/episode_manifest.json` -- `results/omni_finetune/dataset_manifest.json` -- `results/omni_finetune/training_metadata.json` -- `results/omni_finetune/metrics.json` -- `results/omni_finetune/progress.jsonl` -- `results/omni_finetune/RUN_REPORT.md` -- `results/omni_finetune/DATA_BLOCKER_REPORT.md` -- `results/omni_finetune/A100_HF_RELAY_STATUS.md` - -Use this gate before scheduling any 32-episode full fine-tune run. +- access_status: `results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md` -For the A100 Hugging Face relay, the 32-episode pilot should use stratified -selection, not the first 32 paths in repository order. The current relay script -scans 64 top-level session UUIDs, filters for complete leaf episodes, excludes -`visualization.rrd`, applies a `0.25 GB` minimum episode size, and selects 32 -episodes from 32 different session UUIDs. This is still a pilot subset, but it -is materially better for generalization checks than adjacent episodes from the -same recording session. +Use this gate before scheduling any 32-episode full fine-tune run. The pilot +should use stratified selection, not the first 32 paths in repository order. +The current selection plan scans 64 top-level session UUIDs, filters for +complete leaf episodes, excludes `visualization.rrd`, applies a `0.25 GB` +minimum episode size, and selects 32 episodes from 32 different session UUIDs. ### Uploading the pilot Qwen3-Omni LoRA @@ -618,7 +537,7 @@ A prepared upload package is available at `results/omni_finetune/hf_upload`. ```bash python3 scripts/omni/upload_qwen3_omni_lora_to_hf.py \ - --repo-id cy0307/ropedia-qwen3-omni-lora-smoke \ + --repo-id cy0307/ropedia-qwen3-omni-lora-readiness \ --source-dir results/omni_finetune/hf_upload \ --message "Upload Xperience-10M Qwen3-Omni LoRA pilot" ``` diff --git a/QUALITY_GATES.md b/QUALITY_GATES.md index f6f1012be5fefb81e774a9df3e79e24dbf7e15ff..3727fc0220f51d64b1544b5f07cf54bec2cee311 100644 --- a/QUALITY_GATES.md +++ b/QUALITY_GATES.md @@ -12,7 +12,7 @@ These gates validate public packaging, claim boundaries, mirror parity, and webs | Gate | Command | Report | Current report status | Blocks publication if | | --- | --- | --- | --- | --- | -| Scope claims guard | `python scripts/validate_scope_claims.py` | `docs/data/scope_claims_audit.json` | `pass` | Historical 32ep smoke/provenance strings are presented as real 32-episode metrics. | +| Scope claims guard | `python scripts/validate_scope_claims.py` | `docs/data/scope_claims_audit.json` | `pass` | Historical 32ep readiness/provenance strings are presented as real 32-episode metrics. | | Source alignment audit | `python scripts/validate_source_alignment.py` | `docs/data/source_alignment_audit.json` | `pass` | Official full-dataset facts, sample-card facts, API-listing caveats, or public-card boundary markers are missing or inconsistent. | | Website integrity | `python scripts/validate_website_integrity.py` | `docs/data/website_integrity.json` | `pass` | Local links, anchors, JSON bundles, or referenced image assets are missing or invalid. | | Evaluation protocol | `python scripts/build_evaluation_protocol.py` | `docs/data/evaluation_protocol.json` | `pass` | Windowing, split policy, leakage controls, task metrics, or unsupported interpretations are not explicit. | @@ -29,7 +29,7 @@ These gates validate public packaging, claim boundaries, mirror parity, and webs | --- | --- | --- | | Live publication verifier | `python scripts/verify_live_publication.py` | live GitHub Pages, GitHub raw, HF Space, artifact dataset, and model mirrors match the current release assets | | GitHub Pages deployment | `gh run list --repo ChaoYue0307/ropedia-xperience-10m-task-suite --limit 5` | latest pages-build-deployment run succeeds | -| Rendered browser smoke | `Browser/Playwright page identity, nonblank render, console health, and one local interaction` | no relevant console warnings/errors and target links work | +| Rendered browser check | `Browser/Playwright page identity, nonblank render, console health, and one local interaction` | no relevant console warnings/errors and target links work | ## Rerun Order diff --git a/README.md b/README.md index 52294ade53337af796ca1080889951514119e17e..410128fab24b2be872ce954bb8c981b3db1b5ff1 100644 --- a/README.md +++ b/README.md @@ -30,13 +30,10 @@ This dataset repo contains the derived evidence layer for the public Xperience-1 ![12-task infographic](assets/task_suite_infographic.png?v=xperience10m-taskfirst-v12-modality-xl) -The dashboard assets follow a Ropedia-inspired visual system: dark 4D-world -canvas, lime-green accents, point-cloud texture, thin green cards, and -research-grade typography, while all labels and metrics are script-generated -from committed result files. -The project logo is a ChatGPT-image-generated X-shaped multimodal camera mark, -then deterministically packaged into favicon, header, README, Hugging Face -card, and social-preview assets by `scripts/build_brand_assets.py`. +The logo, figures, cards, and website assets are packaged together so the +artifact repo reads as a coherent Xperience-10M multimodal task suite. Labels, +dimensions, and metrics are generated from committed result files rather than +hand-edited presentation copy. The Space starts with a task-first 12-task map, then includes a native responsive modality atlas backed by @@ -79,11 +76,10 @@ the favicon, header, README/HF cards, app icon, and social preview. It does **not** contain raw Xperience-10M videos or raw `annotation.hdf5`. Download raw data only from the official Ropedia / Hugging Face sources and follow their terms. Current scale-up status: the full `ropedia-ai/xperience-10m` Hugging Face -dataset is still gated for this account. The A100 relay has been configured to -poll access, download a 32-episode stratified pilot subset after approval, -validate it, transfer it to H20, and run the readiness gate. Until that -completes, the committed Qwen3-Omni artifacts remain smoke/debug evidence, not -real 32-episode held-out metrics. +dataset is still gated for this account. The multi-episode workflow is prepared +to select, download, validate, and stage a 32-episode held-out pilot after +access approval. Until that completes, the committed Qwen3-Omni artifacts +remain readiness evidence, not real 32-episode held-out metrics. ## Why This Repo Exists @@ -98,7 +94,7 @@ This is the reviewable half of the project. You can inspect the task outputs, co | 3 | How do I reproduce it? | `REPRODUCIBILITY.md`, `docs/data/reproducibility_matrix.json`, `notes/reproducibility_audit.md` | | 4 | What is one model input? | `results/episode_task_suite/windows.csv`, `results/episode_task_suite/feature_manifest.json`, `results/episode_task_suite/available_modalities.json` | | 5 | Are the task results backed by files? | `results/episode_task_suite/summary_report.json`, `results/episode_task_suite/neural_mlp/`, `docs/data/summary_metrics.json` | -| 6 | What is still pending? | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/A100_HF_RELAY_STATUS.md`, `scripts/omni/discover_xperience10m_sources.py` | +| 6 | What is still pending? | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md`, `scripts/omni/discover_xperience10m_sources.py` | Human-readable artifact guide: `ARTIFACT_GUIDE.md`. Reviewer scorecard: `REVIEWER_SCORECARD.md` and `docs/data/reviewer_scorecard.json`. @@ -122,7 +118,7 @@ Source-of-truth brand asset index: `docs/data/brand_assets.json`. | 12-task suite | per-task `metrics.json`, predictions, confusion matrices | chronological single-episode split | | Neural heads | `results/episode_task_suite/neural_mlp/` | compact MLP heads, not a foundation model | | Research directions | `research_direction_taxonomy.json`, extension probe results | direct/proxy/diagnostic evidence, not full solutions | -| Qwen3-Omni | `DATA_BLOCKER_REPORT.md`, `A100_HF_RELAY_STATUS.md` | smoke-only until 32 valid episodes are available | +| Qwen3-Omni | `DATA_BLOCKER_REPORT.md`, `MULTI_EPISODE_ACCESS_STATUS.md` | readiness-only until 32 valid episodes are available | | Scope claims guard | `docs/data/scope_claims_audit.json`, `scripts/validate_scope_claims.py` | historical `32ep` path strings are provenance, not 32-episode results | | Mirror parity | `docs/data/mirror_parity.json`, `scripts/validate_mirror_parity.py` | prepared repo/HF mirrors carry matching critical data, figures, website HTML, and validator files | | Publication hygiene | `docs/data/publication_audit.json`, `scripts/validate_publication_package.py` | public files/HF bundles only, with public-card freshness checks | @@ -156,7 +152,7 @@ Source-of-truth brand asset index: `docs/data/brand_assets.json`. - `FIGURE_INDEX.md` and `docs/data/figure_index.json`: visual evidence index for public figures, charts, thumbnails, dimensions, hashes, and source scripts - `docs/data/artifact_index.json`: source-of-truth proof-artifact catalog with stable-file hashes - `docs/data/mirror_parity.json`: prepared Space/artifact/model mirror parity check, including critical website HTML -- `docs/data/scope_claims_audit.json`: machine-readable guard against overclaiming historical `32ep` smoke-run identifiers +- `docs/data/scope_claims_audit.json`: machine-readable guard against overclaiming historical `32ep` readiness/provenance identifiers - `docs/data/publication_audit.json`: machine-readable publication hygiene and public-card freshness check - `docs/data/website_integrity.json`: machine-readable website local-reference integrity check - `QUALITY_GATES.md` and `docs/data/quality_gates.json`: reviewer-facing and machine-readable release gates @@ -173,7 +169,7 @@ Source-of-truth brand asset index: `docs/data/brand_assets.json`. - `scripts/export_modality_atlas_assets.py`: regenerates the responsive modality-card thumbnails and manifest from the local public sample - `scripts/build_artifact_index.py`: source-of-truth artifact-index builder - `scripts/validate_mirror_parity.py`: prepared mirror parity validator -- `scripts/validate_scope_claims.py`: validates the Qwen3-Omni smoke/result claim boundary +- `scripts/validate_scope_claims.py`: validates the Qwen3-Omni readiness/result claim boundary - `scripts/validate_publication_package.py`: public bundle validator - `scripts/validate_website_integrity.py`: website local-reference validator - `notes/*.md`: interpretation and reproducibility notes diff --git a/REPRODUCIBILITY.md b/REPRODUCIBILITY.md index 3686105208d77c417532fc9213e3bb0845905481..f4bc6131b0ff2ff235a72ea3b09c38858370dbdf 100644 --- a/REPRODUCIBILITY.md +++ b/REPRODUCIBILITY.md @@ -40,7 +40,8 @@ hf download ropedia-ai/xperience-10m-sample \ --local-dir data/sample/xperience-10m-sample ``` -On mainland-China servers, use the included ModelScope helper: +If Hugging Face access is unavailable in your environment, use the included +ModelScope helper: ```bash python scripts/omni/download_sample_modelscope.py \ @@ -136,4 +137,4 @@ Before interpreting any Qwen3-Omni result, read [`docs/data/scope_claims_audit.json`](docs/data/scope_claims_audit.json), [`results/omni_finetune/DATA_BLOCKER_REPORT.md`](results/omni_finetune/DATA_BLOCKER_REPORT.md) and -[`results/omni_finetune/A100_HF_RELAY_STATUS.md`](results/omni_finetune/A100_HF_RELAY_STATUS.md). +[`results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md`](results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md). diff --git a/REVIEWER_SCORECARD.md b/REVIEWER_SCORECARD.md index 3d550ac429684d4dd6a065dba78d46ce9eb1b1da..567370bee404a11363692cc2d5936e3debc9e6d0 100644 --- a/REVIEWER_SCORECARD.md +++ b/REVIEWER_SCORECARD.md @@ -15,7 +15,7 @@ verified only when committed artifacts and validation reports support it. | Website and HF mirrors | Verified | `docs/data/website_integrity.json`, `docs/data/mirror_parity.json`, `docs/data/live_publication_status.json` | Local website links/assets pass, prepared mirrors match, and public GitHub/HF URLs have been checked after upload. | | Publication hygiene | Verified | `docs/data/publication_audit.json`, `QUALITY_GATES.md`, `docs/data/quality_gates.json` | Public bundles are checked for raw-data exclusion, cache exclusion, heavy-archive exclusion, token-string hygiene, and stale presentation copy. | | Reproducibility | Verified for the public sample | `REPRODUCIBILITY.md`, `docs/data/reproducibility_matrix.json`, `notes/reproducibility_audit.md` | The public sample workflow has explicit commands, expected outputs, and exact-match audit evidence. | -| Qwen3-Omni fine-tuning | Data-gated, not a model-quality claim | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/A100_HF_RELAY_STATUS.md` | The 32-episode LoRA pilot is prepared, but no real held-out 32-episode result is claimed until gated data access, manifest construction, training, and held-out evaluation pass. | +| Qwen3-Omni fine-tuning | Data-gated, not a model-quality claim | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md` | The 32-episode LoRA pilot is prepared, but no real held-out 32-episode result is claimed until gated data access, manifest construction, training, and held-out evaluation pass. | | Raw Xperience-10M redistribution | Not included | `DATA_NOTICE.md`, `docs/data/publication_audit.json` | Raw MP4, HDF5, RRD files, private gated data, and full Qwen weights are intentionally excluded. | ## Fast Reviewer Route diff --git a/docs/data/artifact_index.json b/docs/data/artifact_index.json index a11f8bbbb2e3b46070b02fb4bac68c27c4441a7c..6eba91207183bee62aeeee2fae5b352015cf1244 100644 --- a/docs/data/artifact_index.json +++ b/docs/data/artifact_index.json @@ -1,6 +1,6 @@ { "title": "Ropedia Xperience-10M Task Suite Artifact Index", - "generated_at_utc": "2026-06-01T13:19:42+00:00", + "generated_at_utc": "2026-06-01T14:16:54+00:00", "status": "pass", "artifact_count": 49, "missing": [], @@ -39,8 +39,8 @@ "surface": "repo_hf", "proves": "Gives a compact verified/data-gated/not-redistributed decision table for first-pass reviewers.", "exists": true, - "bytes": 4360, - "sha256": "2fe1a0c808eb5906e0942cf1bb573c086ba82523cbda8f778cfa8ac0e582fea4" + "bytes": 4367, + "sha256": "01c3fcd654db595d94b50bc4b68ff1ea9fb13789261dea04dbaba1caa44d5027" }, { "id": "reviewer_scorecard_json", @@ -50,8 +50,8 @@ "surface": "website_hf", "proves": "Machine-readable copy of the current reviewer scorecard for website and HF mirrors.", "exists": true, - "bytes": 6107, - "sha256": "0c368033fa330df11afc1e85010163ba949e108c3e1b3808382d79465a068443" + "bytes": 6114, + "sha256": "6288bddda015e07c0144bffa827c0849858feae5600086287037c005b87a4197" }, { "id": "evidence_contract", @@ -59,10 +59,10 @@ "path": "EVIDENCE_CONTRACT.md", "kind": "claim_boundary", "surface": "repo", - "proves": "Defines what is verified, what is smoke-only, and what must not be inferred.", + "proves": "Defines what is verified, what is readiness-only, and what must not be inferred.", "exists": true, - "bytes": 9701, - "sha256": "7c78f162e48512ae525c02030f20e667fff60d417ba0d4faee82e733ae211f36" + "bytes": 9772, + "sha256": "89f79ed4089e3797338be4711c20e4f31a1dc03e8371fc92faf6ce2fe0c1f355" }, { "id": "reviewer_packet", @@ -72,8 +72,8 @@ "surface": "website_hf", "proves": "Gives a short audit path with scope status and public surfaces.", "exists": true, - "bytes": 7386, - "sha256": "212e98170c90d14beed49c4a6374c8ca9ce07093281a85375e33b60beb2d70fb" + "bytes": 7401, + "sha256": "fefadd40a7ea989dd6d3ff8e4122fc88efc2299f6802a262c24ff1efae8078b9" }, { "id": "artifact_guide", @@ -83,8 +83,8 @@ "surface": "repo_hf", "proves": "Gives the human-readable map from proof boundary to data, tasks, platform mirrors, and scale-up status.", "exists": true, - "bytes": 12376, - "sha256": "6b5e645fdd125dd83b5783f0afce7ec9c353c16fb51bc844d1db701bdbb2662b" + "bytes": 12444, + "sha256": "209eb82a47e66525585b882382412ec01ce38e8561a843430f8c530e7dbf6acd" }, { "id": "official_dataset_card_alignment", @@ -149,8 +149,8 @@ "surface": "repo_hf", "proves": "Defines the window unit, chronological split, task metrics, leakage controls, and unsupported interpretations.", "exists": true, - "bytes": 5844, - "sha256": "6c82f0d6806d2e929f568fec32fcd4820218c59ec6b30bbefe1a4ec1e1aa1b54" + "bytes": 5855, + "sha256": "8fd0836dbe0f99306db0214d53ed56de603734b9ea1b7e6d44b25dbf2a37d168" }, { "id": "evaluation_protocol_json", @@ -160,8 +160,8 @@ "surface": "website_hf", "proves": "Machine-readable protocol generated from committed task metrics for website and HF mirrors.", "exists": true, - "bytes": 13575, - "sha256": "d1cd2724820e8b1f13e0eee70f9d3d805f105d8d8e2ccc638aca1486a6a6e21a" + "bytes": 13586, + "sha256": "01749a94f31059c52adc74cfa36e2a256c43a316a2f318fa4177df3fc2f5a5b4" }, { "id": "evaluation_protocol_builder", @@ -171,8 +171,8 @@ "surface": "repo_hf", "proves": "Regenerates the protocol from committed summary metrics and task artifacts.", "exists": true, - "bytes": 16084, - "sha256": "a859e43fc95b2bce4b85a758050968e504f1dc7870d4b04e230302aec2b74724" + "bytes": 16102, + "sha256": "0781265b37af226432d93b25c18b6278484ba7d6d78e3c56991aaaf78bb0ba76" }, { "id": "figure_index", @@ -215,8 +215,8 @@ "surface": "website_hf", "proves": "Machine-readable manifest for the ChatGPT-image-generated logo, favicon, social card, dimensions, hashes, and usage roles.", "exists": true, - "bytes": 3921, - "sha256": "b284b2dcfa5073d95c1e9eacc97b9bbb21ac0293f93b8e4e3c0cedb5e994ee6f" + "bytes": 3904, + "sha256": "b822ab3af24f6d3f2e041bacf6a1caa59fbd9dbff013c8b21df8f3ef2414c865" }, { "id": "brand_logo_social_card", @@ -237,8 +237,8 @@ "surface": "repo_hf", "proves": "Regenerates logo derivatives, favicon variants, app icons, and the Open Graph social card from the generated logo mark.", "exists": true, - "bytes": 9395, - "sha256": "b9976b12d95ca720b92ee2f02b1faecdc8ef9eea7eb1dea343ee0188fec41ba7" + "bytes": 9378, + "sha256": "90836fea66545f6f90b4a860655efb7263796a6fc86cb243793dd63d459b3a79" }, { "id": "quality_gates", @@ -248,8 +248,8 @@ "surface": "repo_hf", "proves": "Lists the automated and post-publish gates required before presenting a release as current.", "exists": true, - "bytes": 3936, - "sha256": "7e095dc7dd0d98de2913dcad95e23726ee205c81b4a8a277c81e0151df8f9839" + "bytes": 3940, + "sha256": "8979dae77e24bfbea691070637d73540611e16742f5e1f99e834b577ecd72b63" }, { "id": "quality_gate_manifest", @@ -293,8 +293,8 @@ "surface": "repo_hf", "proves": "Defines public reproduction commands, expected outputs, and non-reproducible scale-up boundaries.", "exists": true, - "bytes": 6151, - "sha256": "f1f340ea58a61aeea1ae0cdaaeee250f0458ea616c59c1b39d018c941c18a08b" + "bytes": 6197, + "sha256": "f65169e927c17ad2cb8991c05bdec57b3878c9ed0008aa6e978b675876c46ffc" }, { "id": "reproducibility_matrix", @@ -315,8 +315,8 @@ "surface": "repo_hf", "proves": "Generates the selective proof-artifact catalog from local files.", "exists": true, - "bytes": 18273, - "sha256": "a94e0202c08b4d7d9d5ca57e772b5d15d7a7c698ce0db8f8909bd669e2b0c905" + "bytes": 18358, + "sha256": "180023c3636c67f1a5ca1a1a9afed349011e190c673722a7adf0eb1e7fe1992f" }, { "id": "publication_audit", @@ -327,7 +327,7 @@ "volatile": true, "proves": "Confirms public bundles pass raw-data, cache, archive, and token-string checks.", "exists": true, - "bytes": 6726, + "bytes": 6733, "hash_policy": "existence_and_size_only" }, { @@ -339,7 +339,7 @@ "volatile": true, "proves": "Confirms historical 32ep path strings are not presented as real 32-episode results.", "exists": true, - "bytes": 19964, + "bytes": 20089, "hash_policy": "existence_and_size_only" }, { @@ -396,8 +396,8 @@ "surface": "website_hf", "proves": "Mirrors task metrics for the static dashboard.", "exists": true, - "bytes": 25075, - "sha256": "4daac229d7cad0180009041369d5eb8d6ea97b4889a7f8cbad43c2657b131145" + "bytes": 25088, + "sha256": "6923e492c8b2814b04aea4bdcacef98da94d98dc471ba424a7e977c736dd623f" }, { "id": "feature_manifest", @@ -539,19 +539,19 @@ "surface": "repo_hf", "proves": "Documents why no 32-episode Qwen3-Omni result is claimed yet.", "exists": true, - "bytes": 803, - "sha256": "510d92ba0b1a72bcbe66e6b2e1c1ec325f5e904ed29254a07b1e78a0e5ce3cd3" + "bytes": 813, + "sha256": "561255372fa151fda72b8791746be68c825a320a75264b6131acb743ae381192" }, { - "id": "a100_relay_status", - "title": "A100 relay status", - "path": "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "id": "multi_episode_access_status", + "title": "Multi-episode access status", + "path": "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", "kind": "scaleup_status", "surface": "repo_hf", - "proves": "Documents the pending A100-to-H20 data relay and 32-session pilot selection.", + "proves": "Documents the public multi-episode access boundary and 32-episode pilot selection without exposing private infrastructure details.", "exists": true, - "bytes": 2076, - "sha256": "4d82faff5eb050434a917bc36c0325b972f629952b2e01fe0e33f2647ceede53" + "bytes": 1362, + "sha256": "fd4488d1f21d084b9a2bdf7e769264fe306e48643ff3ef17578f329cfccd5d3c" }, { "id": "citation", diff --git a/docs/data/brand_assets.json b/docs/data/brand_assets.json index 8261373a865f2c0f172eed47386c9ddfae6b68ee..031c7f171262a7449754aa51b083c75d23a39657 100644 --- a/docs/data/brand_assets.json +++ b/docs/data/brand_assets.json @@ -1,11 +1,11 @@ { "title": "Ropedia Xperience-10M Brand Assets", "status": "pass", - "generated_at_utc": "2026-06-01T13:12:30+00:00", + "generated_at_utc": "2026-06-01T14:15:58+00:00", "source": { "path": "docs/assets/brand/xperience10m-logo-mark.png", "kind": "ChatGPT-image-generated logo mark with chroma-key background removed locally", - "prompt_summary": "X-shaped multimodal camera mark with Ropedia-inspired near-black, lime, cyan, trajectory, and point-cloud styling." + "prompt_summary": "X-shaped multimodal camera mark with near-black, lime, cyan, trajectory, and point-cloud styling." }, "assets": [ { diff --git a/docs/data/evaluation_protocol.json b/docs/data/evaluation_protocol.json index 501c2c5bf04abb33c5736debd7c3af8414e9c730..2c0b7deda8052497345dabe9706b6c9a3ecc0747 100644 --- a/docs/data/evaluation_protocol.json +++ b/docs/data/evaluation_protocol.json @@ -2,7 +2,7 @@ "title": "Ropedia Xperience-10M Task Suite Evaluation Protocol", "status": "pass", "version": "2026-06-01", - "generated_at_utc": "2026-06-01T13:12:55+00:00", + "generated_at_utc": "2026-06-01T14:16:26+00:00", "source_files": [ "docs/data/summary_metrics.json", "results/episode_task_suite/summary_report.json", @@ -305,7 +305,7 @@ "unsupported_interpretations": [ "Do not infer cross-episode generalization from this single public sample.", "Do not treat feature-vector reconstruction as pixel depth, mesh, NeRF, or Gaussian reconstruction.", - "Do not treat Qwen3-Omni smoke artifacts as a real 32-episode fine-tune.", + "Do not treat Qwen3-Omni readiness artifacts as a real 32-episode fine-tune.", "Do not infer audio-visual learning from the current baseline vector because audio is not featurized." ], "scale_up_gate": { @@ -318,7 +318,7 @@ "current_status": "prepared but data-gated", "evidence": [ "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md" + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md" ] } } diff --git a/docs/data/evidence_contract.json b/docs/data/evidence_contract.json index 618f94c0a07af6e740e5945e591b0716e8e19945..54d52095fffd28e61d22095e73e16246a50df0c7 100644 --- a/docs/data/evidence_contract.json +++ b/docs/data/evidence_contract.json @@ -142,9 +142,9 @@ "boundary": "single-episode probes, not full research-direction solutions" }, { - "id": "qwen3_omni_smoke", - "claim": "Qwen3-Omni infrastructure has passed technical smoke checks.", - "status": "smoke_only", + "id": "qwen3_omni_readiness", + "claim": "Qwen3-Omni infrastructure has passed technical readiness checks.", + "status": "readiness_only", "evidence": [ "results/omni_finetune/RUN_REPORT.md", "results/omni_finetune/dataset_manifest.json", @@ -158,7 +158,7 @@ "status": "blocked_by_data_access", "evidence": [ "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", "results/omni_finetune/source_discovery.json" ], "boundary": "no 32-episode metric should be claimed until the gate passes" @@ -171,7 +171,7 @@ "scripts/validate_scope_claims.py", "docs/data/scope_claims_audit.json" ], - "boundary": "old run/path identifiers are classified as smoke-artifact provenance and fail validation if public presentation claims real 32-episode metrics" + "boundary": "old run/path identifiers are classified as readiness-artifact provenance and fail validation if public presentation claims real 32-episode metrics" }, { "id": "mirror_parity", diff --git a/docs/data/mirror_parity.json b/docs/data/mirror_parity.json index fcf944a823bcedb97d8fa096ac96f92a931359ab..4fb4176ea4999ef657f6a66e5b6aa7ed2227a0ce 100644 --- a/docs/data/mirror_parity.json +++ b/docs/data/mirror_parity.json @@ -1,6 +1,6 @@ { "status": "pass", - "generated_at_utc": "2026-06-01T13:32:15+00:00", + "generated_at_utc": "2026-06-01T14:32:44+00:00", "hf_root": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish", "summary": { "group_count": 59, @@ -36,27 +36,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/artifact_index.json", "exists": true, - "bytes": 21969, - "sha256": "489cf9ec3b09c3f876927945a947f4f093db5e80b5b2df4c5d728586457e742d" + "bytes": 22054, + "sha256": "1a37c2526c6a63344175ac4be212276c53cc0f654b99df7a875bf6f8d433a413" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/artifact_index.json", "exists": true, - "bytes": 21969, - "sha256": "489cf9ec3b09c3f876927945a947f4f093db5e80b5b2df4c5d728586457e742d" + "bytes": 22054, + "sha256": "1a37c2526c6a63344175ac4be212276c53cc0f654b99df7a875bf6f8d433a413" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/artifact_index.json", "exists": true, - "bytes": 21969, - "sha256": "489cf9ec3b09c3f876927945a947f4f093db5e80b5b2df4c5d728586457e742d" + "bytes": 22054, + "sha256": "1a37c2526c6a63344175ac4be212276c53cc0f654b99df7a875bf6f8d433a413" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/artifact_index.json", "exists": true, - "bytes": 21969, - "sha256": "489cf9ec3b09c3f876927945a947f4f093db5e80b5b2df4c5d728586457e742d" + "bytes": 22054, + "sha256": "1a37c2526c6a63344175ac4be212276c53cc0f654b99df7a875bf6f8d433a413" } }, "failures": [] @@ -67,27 +67,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/brand_assets.json", "exists": true, - "bytes": 3921, - "sha256": "b284b2dcfa5073d95c1e9eacc97b9bbb21ac0293f93b8e4e3c0cedb5e994ee6f" + "bytes": 3904, + "sha256": "b822ab3af24f6d3f2e041bacf6a1caa59fbd9dbff013c8b21df8f3ef2414c865" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/brand_assets.json", "exists": true, - "bytes": 3921, - "sha256": "b284b2dcfa5073d95c1e9eacc97b9bbb21ac0293f93b8e4e3c0cedb5e994ee6f" + "bytes": 3904, + "sha256": "b822ab3af24f6d3f2e041bacf6a1caa59fbd9dbff013c8b21df8f3ef2414c865" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/brand_assets.json", "exists": true, - "bytes": 3921, - "sha256": "b284b2dcfa5073d95c1e9eacc97b9bbb21ac0293f93b8e4e3c0cedb5e994ee6f" + "bytes": 3904, + "sha256": "b822ab3af24f6d3f2e041bacf6a1caa59fbd9dbff013c8b21df8f3ef2414c865" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/brand_assets.json", "exists": true, - "bytes": 3921, - "sha256": "b284b2dcfa5073d95c1e9eacc97b9bbb21ac0293f93b8e4e3c0cedb5e994ee6f" + "bytes": 3904, + "sha256": "b822ab3af24f6d3f2e041bacf6a1caa59fbd9dbff013c8b21df8f3ef2414c865" } }, "failures": [] @@ -98,27 +98,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/evidence_contract.json", "exists": true, - "bytes": 10906, - "sha256": "47c9810067a661a228fa7d3bbe86a4d763cf32caf5eecc13d6112388cca1de42" + "bytes": 10929, + "sha256": "9009dcb98a719854c14c7fa6fe926412d928b1ee45cfbc2cdabfdd4aa62bb501" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/evidence_contract.json", "exists": true, - "bytes": 10906, - "sha256": "47c9810067a661a228fa7d3bbe86a4d763cf32caf5eecc13d6112388cca1de42" + "bytes": 10929, + "sha256": "9009dcb98a719854c14c7fa6fe926412d928b1ee45cfbc2cdabfdd4aa62bb501" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/evidence_contract.json", "exists": true, - "bytes": 10906, - "sha256": "47c9810067a661a228fa7d3bbe86a4d763cf32caf5eecc13d6112388cca1de42" + "bytes": 10929, + "sha256": "9009dcb98a719854c14c7fa6fe926412d928b1ee45cfbc2cdabfdd4aa62bb501" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/evidence_contract.json", "exists": true, - "bytes": 10906, - "sha256": "47c9810067a661a228fa7d3bbe86a4d763cf32caf5eecc13d6112388cca1de42" + "bytes": 10929, + "sha256": "9009dcb98a719854c14c7fa6fe926412d928b1ee45cfbc2cdabfdd4aa62bb501" } }, "failures": [] @@ -129,27 +129,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/evaluation_protocol.json", "exists": true, - "bytes": 13575, - "sha256": "d1cd2724820e8b1f13e0eee70f9d3d805f105d8d8e2ccc638aca1486a6a6e21a" + "bytes": 13586, + "sha256": "01749a94f31059c52adc74cfa36e2a256c43a316a2f318fa4177df3fc2f5a5b4" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/evaluation_protocol.json", "exists": true, - "bytes": 13575, - "sha256": "d1cd2724820e8b1f13e0eee70f9d3d805f105d8d8e2ccc638aca1486a6a6e21a" + "bytes": 13586, + "sha256": "01749a94f31059c52adc74cfa36e2a256c43a316a2f318fa4177df3fc2f5a5b4" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/evaluation_protocol.json", "exists": true, - "bytes": 13575, - "sha256": "d1cd2724820e8b1f13e0eee70f9d3d805f105d8d8e2ccc638aca1486a6a6e21a" + "bytes": 13586, + "sha256": "01749a94f31059c52adc74cfa36e2a256c43a316a2f318fa4177df3fc2f5a5b4" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/evaluation_protocol.json", "exists": true, - "bytes": 13575, - "sha256": "d1cd2724820e8b1f13e0eee70f9d3d805f105d8d8e2ccc638aca1486a6a6e21a" + "bytes": 13586, + "sha256": "01749a94f31059c52adc74cfa36e2a256c43a316a2f318fa4177df3fc2f5a5b4" } }, "failures": [] @@ -284,27 +284,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/publication_audit.json", "exists": true, - "bytes": 6726, - "sha256": "6d53191ee17232e7d88da7b72d8044993cbb7996e9dfff983bcdb8ab206b4341" + "bytes": 6738, + "sha256": "dc2fc110b73a330ad6bec2520f108da245ea0142b33967f9840194fd46491276" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/publication_audit.json", "exists": true, - "bytes": 6726, - "sha256": "6d53191ee17232e7d88da7b72d8044993cbb7996e9dfff983bcdb8ab206b4341" + "bytes": 6738, + "sha256": "dc2fc110b73a330ad6bec2520f108da245ea0142b33967f9840194fd46491276" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/publication_audit.json", "exists": true, - "bytes": 6726, - "sha256": "6d53191ee17232e7d88da7b72d8044993cbb7996e9dfff983bcdb8ab206b4341" + "bytes": 6738, + "sha256": "dc2fc110b73a330ad6bec2520f108da245ea0142b33967f9840194fd46491276" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/publication_audit.json", "exists": true, - "bytes": 6726, - "sha256": "6d53191ee17232e7d88da7b72d8044993cbb7996e9dfff983bcdb8ab206b4341" + "bytes": 6738, + "sha256": "dc2fc110b73a330ad6bec2520f108da245ea0142b33967f9840194fd46491276" } }, "failures": [] @@ -315,27 +315,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/quality_gates.json", "exists": true, - "bytes": 6343, - "sha256": "398c0be6c1894a37a65f16dc52c7497241b40bb65c2f0e7dfffcfa8122ddd1d4" + "bytes": 6351, + "sha256": "5417424b463d37dfe792d4dee57f941f6484433317b2ae6438c85ba49c6cb940" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/quality_gates.json", "exists": true, - "bytes": 6343, - "sha256": "398c0be6c1894a37a65f16dc52c7497241b40bb65c2f0e7dfffcfa8122ddd1d4" + "bytes": 6351, + "sha256": "5417424b463d37dfe792d4dee57f941f6484433317b2ae6438c85ba49c6cb940" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/quality_gates.json", "exists": true, - "bytes": 6343, - "sha256": "398c0be6c1894a37a65f16dc52c7497241b40bb65c2f0e7dfffcfa8122ddd1d4" + "bytes": 6351, + "sha256": "5417424b463d37dfe792d4dee57f941f6484433317b2ae6438c85ba49c6cb940" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/quality_gates.json", "exists": true, - "bytes": 6343, - "sha256": "398c0be6c1894a37a65f16dc52c7497241b40bb65c2f0e7dfffcfa8122ddd1d4" + "bytes": 6351, + "sha256": "5417424b463d37dfe792d4dee57f941f6484433317b2ae6438c85ba49c6cb940" } }, "failures": [] @@ -439,27 +439,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/reviewer_packet.json", "exists": true, - "bytes": 7386, - "sha256": "212e98170c90d14beed49c4a6374c8ca9ce07093281a85375e33b60beb2d70fb" + "bytes": 7401, + "sha256": "fefadd40a7ea989dd6d3ff8e4122fc88efc2299f6802a262c24ff1efae8078b9" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/reviewer_packet.json", "exists": true, - "bytes": 7386, - "sha256": "212e98170c90d14beed49c4a6374c8ca9ce07093281a85375e33b60beb2d70fb" + "bytes": 7401, + "sha256": "fefadd40a7ea989dd6d3ff8e4122fc88efc2299f6802a262c24ff1efae8078b9" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/reviewer_packet.json", "exists": true, - "bytes": 7386, - "sha256": "212e98170c90d14beed49c4a6374c8ca9ce07093281a85375e33b60beb2d70fb" + "bytes": 7401, + "sha256": "fefadd40a7ea989dd6d3ff8e4122fc88efc2299f6802a262c24ff1efae8078b9" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/reviewer_packet.json", "exists": true, - "bytes": 7386, - "sha256": "212e98170c90d14beed49c4a6374c8ca9ce07093281a85375e33b60beb2d70fb" + "bytes": 7401, + "sha256": "fefadd40a7ea989dd6d3ff8e4122fc88efc2299f6802a262c24ff1efae8078b9" } }, "failures": [] @@ -470,27 +470,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/reviewer_scorecard.json", "exists": true, - "bytes": 6107, - "sha256": "0c368033fa330df11afc1e85010163ba949e108c3e1b3808382d79465a068443" + "bytes": 6114, + "sha256": "6288bddda015e07c0144bffa827c0849858feae5600086287037c005b87a4197" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/reviewer_scorecard.json", "exists": true, - "bytes": 6107, - "sha256": "0c368033fa330df11afc1e85010163ba949e108c3e1b3808382d79465a068443" + "bytes": 6114, + "sha256": "6288bddda015e07c0144bffa827c0849858feae5600086287037c005b87a4197" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/reviewer_scorecard.json", "exists": true, - "bytes": 6107, - "sha256": "0c368033fa330df11afc1e85010163ba949e108c3e1b3808382d79465a068443" + "bytes": 6114, + "sha256": "6288bddda015e07c0144bffa827c0849858feae5600086287037c005b87a4197" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/reviewer_scorecard.json", "exists": true, - "bytes": 6107, - "sha256": "0c368033fa330df11afc1e85010163ba949e108c3e1b3808382d79465a068443" + "bytes": 6114, + "sha256": "6288bddda015e07c0144bffa827c0849858feae5600086287037c005b87a4197" } }, "failures": [] @@ -501,27 +501,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/scope_claims_audit.json", "exists": true, - "bytes": 19964, - "sha256": "356566b88c87b7215c4869b5406560b8d85359b1f46bf004935f9edec39f16b5" + "bytes": 20089, + "sha256": "5d4d1b799bcdb59fcda3cb5d4ddce59e99e8033cc675b63594b1db0c556002c6" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/scope_claims_audit.json", "exists": true, - "bytes": 19964, - "sha256": "356566b88c87b7215c4869b5406560b8d85359b1f46bf004935f9edec39f16b5" + "bytes": 20089, + "sha256": "5d4d1b799bcdb59fcda3cb5d4ddce59e99e8033cc675b63594b1db0c556002c6" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/scope_claims_audit.json", "exists": true, - "bytes": 19964, - "sha256": "356566b88c87b7215c4869b5406560b8d85359b1f46bf004935f9edec39f16b5" + "bytes": 20089, + "sha256": "5d4d1b799bcdb59fcda3cb5d4ddce59e99e8033cc675b63594b1db0c556002c6" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/scope_claims_audit.json", "exists": true, - "bytes": 19964, - "sha256": "356566b88c87b7215c4869b5406560b8d85359b1f46bf004935f9edec39f16b5" + "bytes": 20089, + "sha256": "5d4d1b799bcdb59fcda3cb5d4ddce59e99e8033cc675b63594b1db0c556002c6" } }, "failures": [] @@ -533,26 +533,26 @@ "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/source_alignment_audit.json", "exists": true, "bytes": 4425, - "sha256": "b025f11201d0f32f54930f8956d606a6526d2d32b1a7c9cdf5382ddf2bec8fee" + "sha256": "d0544dce6980d03ad73f3ee9ec9e9890de448b28f2a7bb7f7d6c715d7bc67157" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/source_alignment_audit.json", "exists": true, "bytes": 4425, - "sha256": "b025f11201d0f32f54930f8956d606a6526d2d32b1a7c9cdf5382ddf2bec8fee" + "sha256": "d0544dce6980d03ad73f3ee9ec9e9890de448b28f2a7bb7f7d6c715d7bc67157" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/source_alignment_audit.json", "exists": true, "bytes": 4425, - "sha256": "b025f11201d0f32f54930f8956d606a6526d2d32b1a7c9cdf5382ddf2bec8fee" + "sha256": "d0544dce6980d03ad73f3ee9ec9e9890de448b28f2a7bb7f7d6c715d7bc67157" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/source_alignment_audit.json", "exists": true, "bytes": 4425, - "sha256": "b025f11201d0f32f54930f8956d606a6526d2d32b1a7c9cdf5382ddf2bec8fee" + "sha256": "d0544dce6980d03ad73f3ee9ec9e9890de448b28f2a7bb7f7d6c715d7bc67157" } }, "failures": [] @@ -563,27 +563,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/summary_metrics.json", "exists": true, - "bytes": 25075, - "sha256": "4daac229d7cad0180009041369d5eb8d6ea97b4889a7f8cbad43c2657b131145" + "bytes": 25088, + "sha256": "6923e492c8b2814b04aea4bdcacef98da94d98dc471ba424a7e977c736dd623f" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/summary_metrics.json", "exists": true, - "bytes": 25075, - "sha256": "4daac229d7cad0180009041369d5eb8d6ea97b4889a7f8cbad43c2657b131145" + "bytes": 25088, + "sha256": "6923e492c8b2814b04aea4bdcacef98da94d98dc471ba424a7e977c736dd623f" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/summary_metrics.json", "exists": true, - "bytes": 25075, - "sha256": "4daac229d7cad0180009041369d5eb8d6ea97b4889a7f8cbad43c2657b131145" + "bytes": 25088, + "sha256": "6923e492c8b2814b04aea4bdcacef98da94d98dc471ba424a7e977c736dd623f" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/summary_metrics.json", "exists": true, - "bytes": 25075, - "sha256": "4daac229d7cad0180009041369d5eb8d6ea97b4889a7f8cbad43c2657b131145" + "bytes": 25088, + "sha256": "6923e492c8b2814b04aea4bdcacef98da94d98dc471ba424a7e977c736dd623f" } }, "failures": [] @@ -626,26 +626,26 @@ "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/data/website_integrity.json", "exists": true, "bytes": 8813, - "sha256": "db9d245aec43c92d95dd8e0506ad4f7208aec775579df3f7142b59d0c4facc87" + "sha256": "ec2d694a395035f5f47b3367e72f7e7f97ab4300908096a14676800d591687e8" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/data/website_integrity.json", "exists": true, "bytes": 8813, - "sha256": "db9d245aec43c92d95dd8e0506ad4f7208aec775579df3f7142b59d0c4facc87" + "sha256": "ec2d694a395035f5f47b3367e72f7e7f97ab4300908096a14676800d591687e8" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/data/website_integrity.json", "exists": true, "bytes": 8813, - "sha256": "db9d245aec43c92d95dd8e0506ad4f7208aec775579df3f7142b59d0c4facc87" + "sha256": "ec2d694a395035f5f47b3367e72f7e7f97ab4300908096a14676800d591687e8" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/metrics/website_integrity.json", "exists": true, "bytes": 8813, - "sha256": "db9d245aec43c92d95dd8e0506ad4f7208aec775579df3f7142b59d0c4facc87" + "sha256": "ec2d694a395035f5f47b3367e72f7e7f97ab4300908096a14676800d591687e8" } }, "failures": [] @@ -1316,21 +1316,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/build_artifact_index.py", "exists": true, - "bytes": 18273, - "sha256": "a94e0202c08b4d7d9d5ca57e772b5d15d7a7c698ce0db8f8909bd669e2b0c905" + "bytes": 18358, + "sha256": "180023c3636c67f1a5ca1a1a9afed349011e190c673722a7adf0eb1e7fe1992f" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/build_artifact_index.py", "exists": true, - "bytes": 18273, - "sha256": "a94e0202c08b4d7d9d5ca57e772b5d15d7a7c698ce0db8f8909bd669e2b0c905" + "bytes": 18358, + "sha256": "180023c3636c67f1a5ca1a1a9afed349011e190c673722a7adf0eb1e7fe1992f" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/build_artifact_index.py", "exists": true, - "bytes": 18273, - "sha256": "a94e0202c08b4d7d9d5ca57e772b5d15d7a7c698ce0db8f8909bd669e2b0c905" + "bytes": 18358, + "sha256": "180023c3636c67f1a5ca1a1a9afed349011e190c673722a7adf0eb1e7fe1992f" } }, "failures": [] @@ -1341,21 +1341,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/build_brand_assets.py", "exists": true, - "bytes": 9395, - "sha256": "b9976b12d95ca720b92ee2f02b1faecdc8ef9eea7eb1dea343ee0188fec41ba7" + "bytes": 9378, + "sha256": "90836fea66545f6f90b4a860655efb7263796a6fc86cb243793dd63d459b3a79" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/build_brand_assets.py", "exists": true, - "bytes": 9395, - "sha256": "b9976b12d95ca720b92ee2f02b1faecdc8ef9eea7eb1dea343ee0188fec41ba7" + "bytes": 9378, + "sha256": "90836fea66545f6f90b4a860655efb7263796a6fc86cb243793dd63d459b3a79" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/build_brand_assets.py", "exists": true, - "bytes": 9395, - "sha256": "b9976b12d95ca720b92ee2f02b1faecdc8ef9eea7eb1dea343ee0188fec41ba7" + "bytes": 9378, + "sha256": "90836fea66545f6f90b4a860655efb7263796a6fc86cb243793dd63d459b3a79" } }, "failures": [] @@ -1366,21 +1366,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/build_evaluation_protocol.py", "exists": true, - "bytes": 16084, - "sha256": "a859e43fc95b2bce4b85a758050968e504f1dc7870d4b04e230302aec2b74724" + "bytes": 16102, + "sha256": "0781265b37af226432d93b25c18b6278484ba7d6d78e3c56991aaaf78bb0ba76" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/build_evaluation_protocol.py", "exists": true, - "bytes": 16084, - "sha256": "a859e43fc95b2bce4b85a758050968e504f1dc7870d4b04e230302aec2b74724" + "bytes": 16102, + "sha256": "0781265b37af226432d93b25c18b6278484ba7d6d78e3c56991aaaf78bb0ba76" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/build_evaluation_protocol.py", "exists": true, - "bytes": 16084, - "sha256": "a859e43fc95b2bce4b85a758050968e504f1dc7870d4b04e230302aec2b74724" + "bytes": 16102, + "sha256": "0781265b37af226432d93b25c18b6278484ba7d6d78e3c56991aaaf78bb0ba76" } }, "failures": [] @@ -1416,21 +1416,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/build_quality_gates.py", "exists": true, - "bytes": 9800, - "sha256": "e7027b910b1eff67bca47cd64f65df5f0605c74ffc493a02d0223ff66c5744c8" + "bytes": 9808, + "sha256": "edd8cf0ef26091cdbaa1946b3db107e40e04d7329ca6fdbb83b4db2451bdc32b" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/build_quality_gates.py", "exists": true, - "bytes": 9800, - "sha256": "e7027b910b1eff67bca47cd64f65df5f0605c74ffc493a02d0223ff66c5744c8" + "bytes": 9808, + "sha256": "edd8cf0ef26091cdbaa1946b3db107e40e04d7329ca6fdbb83b4db2451bdc32b" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/build_quality_gates.py", "exists": true, - "bytes": 9800, - "sha256": "e7027b910b1eff67bca47cd64f65df5f0605c74ffc493a02d0223ff66c5744c8" + "bytes": 9808, + "sha256": "edd8cf0ef26091cdbaa1946b3db107e40e04d7329ca6fdbb83b4db2451bdc32b" } }, "failures": [] @@ -1491,21 +1491,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/validate_publication_package.py", "exists": true, - "bytes": 15608, - "sha256": "de874d0f343f5daeb6a4ee71b32b1d54f31f1c9191cb70d3b3158b9b660f3b9d" + "bytes": 15615, + "sha256": "b46a46ba90597fdfbf43bae20375bb1d6618fd2c3835a42c770017b572119780" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/validate_publication_package.py", "exists": true, - "bytes": 15608, - "sha256": "de874d0f343f5daeb6a4ee71b32b1d54f31f1c9191cb70d3b3158b9b660f3b9d" + "bytes": 15615, + "sha256": "b46a46ba90597fdfbf43bae20375bb1d6618fd2c3835a42c770017b572119780" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/validate_publication_package.py", "exists": true, - "bytes": 15608, - "sha256": "de874d0f343f5daeb6a4ee71b32b1d54f31f1c9191cb70d3b3158b9b660f3b9d" + "bytes": 15615, + "sha256": "b46a46ba90597fdfbf43bae20375bb1d6618fd2c3835a42c770017b572119780" } }, "failures": [] @@ -1516,21 +1516,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/scripts/validate_scope_claims.py", "exists": true, - "bytes": 11510, - "sha256": "bdceff4d445a12d9a45ab66a7e101e6205af306955c7d23fd3258fc8eed92c3f" + "bytes": 11545, + "sha256": "caa4f56277a89e78c0b1dc0a7e0d310f77c5f569c059e249cd07377146cf2e88" }, "mirrors": { "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/scripts/validate_scope_claims.py", "exists": true, - "bytes": 11510, - "sha256": "bdceff4d445a12d9a45ab66a7e101e6205af306955c7d23fd3258fc8eed92c3f" + "bytes": 11545, + "sha256": "caa4f56277a89e78c0b1dc0a7e0d310f77c5f569c059e249cd07377146cf2e88" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/scripts/validate_scope_claims.py", "exists": true, - "bytes": 11510, - "sha256": "bdceff4d445a12d9a45ab66a7e101e6205af306955c7d23fd3258fc8eed92c3f" + "bytes": 11545, + "sha256": "caa4f56277a89e78c0b1dc0a7e0d310f77c5f569c059e249cd07377146cf2e88" } }, "failures": [] @@ -1641,21 +1641,21 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs/index.html", "exists": true, - "bytes": 106505, - "sha256": "8f68101f8537c2de91b53c7a07a2db453bedf4d7e34f25cde22944e5a35b85cb" + "bytes": 106467, + "sha256": "9bc5053b4e73028263bd9daf5065af75a2dbf9222ebc78d5ad9c30d5e03211e3" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/index.html", "exists": true, - "bytes": 106505, - "sha256": "8f68101f8537c2de91b53c7a07a2db453bedf4d7e34f25cde22944e5a35b85cb" + "bytes": 106467, + "sha256": "9bc5053b4e73028263bd9daf5065af75a2dbf9222ebc78d5ad9c30d5e03211e3" }, "hf_artifacts_docs": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/docs/index.html", "exists": true, - "bytes": 106505, - "sha256": "8f68101f8537c2de91b53c7a07a2db453bedf4d7e34f25cde22944e5a35b85cb" + "bytes": 106467, + "sha256": "9bc5053b4e73028263bd9daf5065af75a2dbf9222ebc78d5ad9c30d5e03211e3" } }, "failures": [] @@ -1691,27 +1691,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/QUALITY_GATES.md", "exists": true, - "bytes": 3936, - "sha256": "7e095dc7dd0d98de2913dcad95e23726ee205c81b4a8a277c81e0151df8f9839" + "bytes": 3940, + "sha256": "8979dae77e24bfbea691070637d73540611e16742f5e1f99e834b577ecd72b63" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/QUALITY_GATES.md", "exists": true, - "bytes": 3936, - "sha256": "7e095dc7dd0d98de2913dcad95e23726ee205c81b4a8a277c81e0151df8f9839" + "bytes": 3940, + "sha256": "8979dae77e24bfbea691070637d73540611e16742f5e1f99e834b577ecd72b63" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/QUALITY_GATES.md", "exists": true, - "bytes": 3936, - "sha256": "7e095dc7dd0d98de2913dcad95e23726ee205c81b4a8a277c81e0151df8f9839" + "bytes": 3940, + "sha256": "8979dae77e24bfbea691070637d73540611e16742f5e1f99e834b577ecd72b63" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/QUALITY_GATES.md", "exists": true, - "bytes": 3936, - "sha256": "7e095dc7dd0d98de2913dcad95e23726ee205c81b4a8a277c81e0151df8f9839" + "bytes": 3940, + "sha256": "8979dae77e24bfbea691070637d73540611e16742f5e1f99e834b577ecd72b63" } }, "failures": [] @@ -1722,27 +1722,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/EVALUATION_PROTOCOL.md", "exists": true, - "bytes": 5844, - "sha256": "6c82f0d6806d2e929f568fec32fcd4820218c59ec6b30bbefe1a4ec1e1aa1b54" + "bytes": 5855, + "sha256": "8fd0836dbe0f99306db0214d53ed56de603734b9ea1b7e6d44b25dbf2a37d168" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/EVALUATION_PROTOCOL.md", "exists": true, - "bytes": 5844, - "sha256": "6c82f0d6806d2e929f568fec32fcd4820218c59ec6b30bbefe1a4ec1e1aa1b54" + "bytes": 5855, + "sha256": "8fd0836dbe0f99306db0214d53ed56de603734b9ea1b7e6d44b25dbf2a37d168" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/EVALUATION_PROTOCOL.md", "exists": true, - "bytes": 5844, - "sha256": "6c82f0d6806d2e929f568fec32fcd4820218c59ec6b30bbefe1a4ec1e1aa1b54" + "bytes": 5855, + "sha256": "8fd0836dbe0f99306db0214d53ed56de603734b9ea1b7e6d44b25dbf2a37d168" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/EVALUATION_PROTOCOL.md", "exists": true, - "bytes": 5844, - "sha256": "6c82f0d6806d2e929f568fec32fcd4820218c59ec6b30bbefe1a4ec1e1aa1b54" + "bytes": 5855, + "sha256": "8fd0836dbe0f99306db0214d53ed56de603734b9ea1b7e6d44b25dbf2a37d168" } }, "failures": [] @@ -1784,27 +1784,27 @@ "local": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/REVIEWER_SCORECARD.md", "exists": true, - "bytes": 4360, - "sha256": "2fe1a0c808eb5906e0942cf1bb573c086ba82523cbda8f778cfa8ac0e582fea4" + "bytes": 4367, + "sha256": "01c3fcd654db595d94b50bc4b68ff1ea9fb13789261dea04dbaba1caa44d5027" }, "mirrors": { "hf_space": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/space/REVIEWER_SCORECARD.md", "exists": true, - "bytes": 4360, - "sha256": "2fe1a0c808eb5906e0942cf1bb573c086ba82523cbda8f778cfa8ac0e582fea4" + "bytes": 4367, + "sha256": "01c3fcd654db595d94b50bc4b68ff1ea9fb13789261dea04dbaba1caa44d5027" }, "hf_artifacts": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts/REVIEWER_SCORECARD.md", "exists": true, - "bytes": 4360, - "sha256": "2fe1a0c808eb5906e0942cf1bb573c086ba82523cbda8f778cfa8ac0e582fea4" + "bytes": 4367, + "sha256": "01c3fcd654db595d94b50bc4b68ff1ea9fb13789261dea04dbaba1caa44d5027" }, "hf_model": { "path": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model/REVIEWER_SCORECARD.md", "exists": true, - "bytes": 4360, - "sha256": "2fe1a0c808eb5906e0942cf1bb573c086ba82523cbda8f778cfa8ac0e582fea4" + "bytes": 4367, + "sha256": "01c3fcd654db595d94b50bc4b68ff1ea9fb13789261dea04dbaba1caa44d5027" } }, "failures": [] diff --git a/docs/data/publication_audit.json b/docs/data/publication_audit.json index 2675a198c0865157022eb444a9d29ebb21742a20..b389ce9b65aff5b33e388caf7f78986df7072290 100644 --- a/docs/data/publication_audit.json +++ b/docs/data/publication_audit.json @@ -1,6 +1,6 @@ { "status": "pass", - "generated_at_utc": "2026-06-01T13:32:15+00:00", + "generated_at_utc": "2026-06-01T14:30:09+00:00", "checks": [ { "name": "required_publication_assets_present", @@ -100,7 +100,7 @@ "results/episode_task_suite/feature_manifest.json": true, "results/episode_task_suite/neural_mlp/timeline_action/metrics.json": true, "results/omni_finetune/DATA_BLOCKER_REPORT.md": true, - "results/omni_finetune/A100_HF_RELAY_STATUS.md": true, + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md": true, "scripts/episode_task_suite.py": true, "scripts/neural_task_models.py": true, "scripts/build_artifact_index.py": true, @@ -183,19 +183,19 @@ "hf_artifact_bundle": { "root": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/artifacts", "exists": true, - "file_count": 303, - "text_file_count": 245, + "file_count": 359, + "text_file_count": 273, "largest_file": { - "path": "results/episode_task_suite/neural_mlp/temporal_order/model.pt", - "bytes": 13406129 + "path": "results/episode_task_suite/modality_reconstruction/predictions.npz", + "bytes": 52601010 }, "violations": [] }, "hf_model_bundle": { "root": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/hf_publish/model", "exists": true, - "file_count": 225, - "text_file_count": 175, + "file_count": 224, + "text_file_count": 174, "largest_file": { "path": "artifacts/episode_task_suite/cross_modal_retrieval/model.npz", "bytes": 41310574 diff --git a/docs/data/quality_gates.json b/docs/data/quality_gates.json index 145c5aa78b6a6de7320deb23dfbdb27e63e94a36..531b85a107f03e7e10f893d407e4f89ee3640acc 100644 --- a/docs/data/quality_gates.json +++ b/docs/data/quality_gates.json @@ -1,7 +1,7 @@ { "title": "Ropedia Xperience-10M Publication Quality Gates", "status": "pass", - "generated_at_utc": "2026-06-01T13:19:42+00:00", + "generated_at_utc": "2026-06-01T14:17:03+00:00", "rule": "Do not present a release as current unless every automated gate passes, then verify live GitHub/HF mirrors after publishing.", "automated_gates": [ { @@ -9,8 +9,8 @@ "title": "Scope claims guard", "command": "python scripts/validate_scope_claims.py", "report": "docs/data/scope_claims_audit.json", - "blocks_if": "Historical 32ep smoke/provenance strings are presented as real 32-episode metrics.", - "proves": "The public narrative does not overclaim the Qwen3-Omni smoke artifacts.", + "blocks_if": "Historical 32ep readiness/provenance strings are presented as real 32-episode metrics.", + "proves": "The public narrative does not overclaim the Qwen3-Omni readiness artifacts.", "current_report": { "exists": true, "status": "pass" @@ -139,8 +139,8 @@ "required_result": "latest pages-build-deployment run succeeds" }, { - "id": "rendered_browser_smoke", - "title": "Rendered browser smoke", + "id": "rendered_browser_check", + "title": "Rendered browser check", "evidence": "Browser/Playwright page identity, nonblank render, console health, and one local interaction", "required_result": "no relevant console warnings/errors and target links work" } diff --git a/docs/data/reviewer_packet.json b/docs/data/reviewer_packet.json index a75761cf60cc6b78e7d5e5b0f877b28a79c6e0b1..8250ff8164121f2e8e1a1d4252700fa84199fd97 100644 --- a/docs/data/reviewer_packet.json +++ b/docs/data/reviewer_packet.json @@ -12,7 +12,7 @@ "raw_xperience10m_data_in_repo": false, "audio_feature_status": "Audio is present in the sample MP4 streams and shown in the figures, but the current baseline feature vector does not include an extracted audio block.", "qwen3_omni_32_episode_claim": false, - "qwen3_omni_status": "Smoke-only until at least 32 valid episodes are available and held-out episode evaluation finishes." + "qwen3_omni_status": "Readiness-only until at least 32 valid episodes are available and held-out episode evaluation finishes." }, "review_path": [ { @@ -108,7 +108,7 @@ "question": "How should this scale beyond one episode?", "primary_artifacts": [ "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", "scripts/omni/discover_xperience10m_sources.py" ], "readout": "The next milestone is a 32-episode held-out-episode Qwen3-Omni LoRA pilot after gated Xperience-10M access is available." @@ -136,7 +136,7 @@ }, "do_not_infer": [ "Do not infer cross-environment generalization from the single public sample episode.", - "Do not treat the Qwen3-Omni smoke run as a 32-episode fine-tune.", + "Do not treat the Qwen3-Omni readiness run as a 32-episode fine-tune.", "Do not treat feature-vector reconstruction as pixel-depth, mesh, NeRF, or Gaussian reconstruction.", "Do not assume raw Xperience-10M data is redistributed in this repo." ] diff --git a/docs/data/reviewer_scorecard.json b/docs/data/reviewer_scorecard.json index 958e591ebf8d944fc336ed81ce56cdc7d3ebb76a..6dd29c02f65da4301a35963601d53d246b81a58a 100644 --- a/docs/data/reviewer_scorecard.json +++ b/docs/data/reviewer_scorecard.json @@ -108,7 +108,7 @@ "status": "data_gated_not_model_quality_claim", "evidence": [ "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md" + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md" ], "readout": "The 32-episode LoRA pilot is prepared, but no real held-out 32-episode result is claimed until gated data access, manifest construction, training, and held-out evaluation pass." }, diff --git a/docs/data/scope_claims_audit.json b/docs/data/scope_claims_audit.json index 78303432265b065d02f8e2ab4ee68ba56395d297..a1b1c911b737254893daf2ddacaa4c56c05e6d08 100644 --- a/docs/data/scope_claims_audit.json +++ b/docs/data/scope_claims_audit.json @@ -1,6 +1,6 @@ { "status": "pass", - "generated_at_utc": "2026-06-01T13:13:35+00:00", + "generated_at_utc": "2026-06-01T14:16:38+00:00", "summary": { "qwen3_omni_32_episode_claim": false, "dataset_manifest_num_episodes": 1, @@ -31,7 +31,7 @@ { "name": "reviewer_packet_forbids_32_episode_inference", "status": "pass", - "detail": "reviewer packet explicitly warns not to treat the smoke run as a 32-episode fine-tune", + "detail": "reviewer packet explicitly warns not to treat the readiness run as a 32-episode fine-tune", "evidence": [ "docs/data/reviewer_packet.json" ] @@ -39,13 +39,13 @@ { "name": "summary_metrics_preserves_omni_claim_boundary", "status": "pass", - "detail": "No real 32-episode fine-tune is claimed until the watcher downloads data, transfers it to H20, and the held-out evaluation runs.", + "detail": "No real 32-episode fine-tune is claimed until gated data is available locally and held-out evaluation runs.", "evidence": [ "docs/data/summary_metrics.json" ] }, { - "name": "omni_dataset_manifest_is_smoke_only", + "name": "omni_dataset_manifest_is_readiness_only", "status": "pass", "detail": "episodes=1, samples=128, split_counts={'train': 128}", "evidence": [ @@ -53,7 +53,7 @@ ] }, { - "name": "omni_training_metadata_is_smoke_only", + "name": "omni_training_metadata_is_readiness_only", "status": "pass", "detail": "train=128, val=0, processes=8", "evidence": [ @@ -88,7 +88,7 @@ ] }, { - "name": "historical_32ep_identifiers_are_confined_to_smoke_artifacts", + "name": "historical_32ep_identifiers_are_confined_to_readiness_artifacts", "status": "pass", "detail": "historical identifiers found in result provenance files=157", "evidence": [ @@ -140,7 +140,7 @@ ], "historical_identifiers": [ { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/HF_UPLOAD.md", "line": 5, "patterns": [ @@ -150,7 +150,7 @@ "example": "- `results/omni_finetune/adapter_lora/` (`xperience10m_qwen3_omni_32ep_lora`)" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/RUN_REPORT.md", "line": 4, "patterns": [ @@ -160,7 +160,7 @@ "example": "- Adapter: `checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora`" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/RUN_REPORT.md", "line": 5, "patterns": [ @@ -170,7 +170,7 @@ "example": "- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl`" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/RUN_REPORT_eval.md", "line": 4, "patterns": [ @@ -180,7 +180,7 @@ "example": "- Adapter: `checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora`" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/RUN_REPORT_eval.md", "line": 5, "patterns": [ @@ -190,7 +190,7 @@ "example": "- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl`" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/RUN_REPORT_lora.md", "line": 4, "patterns": [ @@ -200,7 +200,7 @@ "example": "- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl`" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/config.yaml", "line": 1, "patterns": [ @@ -210,7 +210,7 @@ "example": "run_id: xperience10m_qwen3_omni_32ep_lora" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/config.yaml", "line": 4, "patterns": [ @@ -220,7 +220,7 @@ "example": "dataset_jsonl: results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/config.yaml", "line": 5, "patterns": [ @@ -228,10 +228,10 @@ "xperience10m_qwen3_omni_32ep", "ropedia-episode-task-suite" ], - "example": "checkpoint_dir: /home/cy/Ropedia/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora" + "example": "checkpoint_dir: /path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 1, "patterns": [ @@ -242,7 +242,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:0\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 0, \"end_frame\": 19, \"num_frames\": 20}, \"media\": {\"video_paths\": [{" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 2, "patterns": [ @@ -253,7 +253,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:1\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 20, \"end_frame\": 39, \"num_frames\": 20}, \"media\": {\"video_paths\": [" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 3, "patterns": [ @@ -264,7 +264,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:2\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 40, \"end_frame\": 59, \"num_frames\": 20}, \"media\": {\"video_paths\": [" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 4, "patterns": [ @@ -275,7 +275,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:3\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 60, \"end_frame\": 79, \"num_frames\": 20}, \"media\": {\"video_paths\": [" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 5, "patterns": [ @@ -286,7 +286,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:4\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 80, \"end_frame\": 99, \"num_frames\": 20}, \"media\": {\"video_paths\": [" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 6, "patterns": [ @@ -297,7 +297,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:5\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 100, \"end_frame\": 119, \"num_frames\": 20}, \"media\": {\"video_paths\":" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 7, "patterns": [ @@ -308,7 +308,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:6\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 120, \"end_frame\": 139, \"num_frames\": 20}, \"media\": {\"video_paths\":" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 8, "patterns": [ @@ -319,7 +319,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:7\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 140, \"end_frame\": 159, \"num_frames\": 20}, \"media\": {\"video_paths\":" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 9, "patterns": [ @@ -330,7 +330,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:8\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 160, \"end_frame\": 179, \"num_frames\": 20}, \"media\": {\"video_paths\":" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 10, "patterns": [ @@ -341,7 +341,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:9\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 180, \"end_frame\": 199, \"num_frames\": 20}, \"media\": {\"video_paths\":" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 11, "patterns": [ @@ -352,7 +352,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:10\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 200, \"end_frame\": 219, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 12, "patterns": [ @@ -363,7 +363,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:11\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 220, \"end_frame\": 239, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 13, "patterns": [ @@ -374,7 +374,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:12\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 240, \"end_frame\": 259, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 14, "patterns": [ @@ -385,7 +385,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:13\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 260, \"end_frame\": 279, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 15, "patterns": [ @@ -396,7 +396,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:14\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 280, \"end_frame\": 299, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 16, "patterns": [ @@ -407,7 +407,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:15\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 300, \"end_frame\": 319, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 17, "patterns": [ @@ -418,7 +418,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:40\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 800, \"end_frame\": 819, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 18, "patterns": [ @@ -429,7 +429,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:41\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 820, \"end_frame\": 839, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 19, "patterns": [ @@ -440,7 +440,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:42\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 840, \"end_frame\": 859, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 20, "patterns": [ @@ -451,7 +451,7 @@ "example": "{\"id\": \"xperience-10m-sample:qa:43\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 860, \"end_frame\": 879, \"num_frames\": 20}, \"media\": {\"video_paths\"" }, { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": "results/omni_finetune/dataset.jsonl", "line": 21, "patterns": [ diff --git a/docs/data/source_alignment_audit.json b/docs/data/source_alignment_audit.json index 40f99f6af7328f5957a75614eeb42efac77b530f..706751fa70bbcd62eabb39b6583109f68a5f1866 100644 --- a/docs/data/source_alignment_audit.json +++ b/docs/data/source_alignment_audit.json @@ -1,7 +1,7 @@ { "title": "Ropedia Xperience-10M Source Alignment Audit", "status": "pass", - "generated_at_utc": "2026-06-01T13:12:55+00:00", + "generated_at_utc": "2026-06-01T14:26:41+00:00", "alignment_json": "docs/data/xperience10m_dataset_card_alignment.json", "alignment_summary": { "full_dataset_repo": "ropedia-ai/xperience-10m", diff --git a/docs/data/summary_metrics.json b/docs/data/summary_metrics.json index 2734ee0bbd712326b5aa84af39be2d643211694a..9811a329097464fd2bbc56b4ea361f6c98d7b9e2 100644 --- a/docs/data/summary_metrics.json +++ b/docs/data/summary_metrics.json @@ -2,8 +2,8 @@ "omni_relay": { "status": "pending_huggingface_gated_access", "dataset": "ropedia-ai/xperience-10m", - "relay_server": "ANGEL-A100-80Gx4", - "training_server": "ANGEL-H20-96GX8", + "staging": "prepared_generic_host_to_host_transfer", + "training_target": "external_multi_gpu_training_host", "selection_strategy": "stratified_round_robin_by_top_level_session", "target_episodes": 32, "selected_sessions": 32, @@ -14,7 +14,7 @@ "visualization.rrd" ], "blocker": "Hugging Face returns 403 pending review for the full Xperience-10M gated dataset.", - "claim_boundary": "No real 32-episode fine-tune is claimed until the watcher downloads data, transfers it to H20, and the held-out evaluation runs." + "claim_boundary": "No real 32-episode fine-tune is claimed until gated data is available locally and held-out evaluation runs." }, "models": { "motion_action": { diff --git a/docs/data/website_integrity.json b/docs/data/website_integrity.json index 9e59a1b3309a7c243462372665b95498b9299c62..7d921e7def499cb812ef654dda022ffc0605a88f 100644 --- a/docs/data/website_integrity.json +++ b/docs/data/website_integrity.json @@ -1,6 +1,6 @@ { "status": "pass", - "generated_at_utc": "2026-06-01T13:12:56+00:00", + "generated_at_utc": "2026-06-01T14:22:00+00:00", "docs_root": "/Users/chaoyue/Documents/Codex/2026-05-29/i-am-learning-this-dataset-https/working_repo_copy/docs", "site_base": "/ropedia-xperience-10m-task-suite/", "summary": { @@ -25,7 +25,7 @@ "status": "pass", "reason": "The reviewer scorecard should appear before the deeper evidence ledger.", "scorecard_index": 37042, - "evidence_index": 43678 + "evidence_index": 43688 }, { "name": "reviewer_scorecard_links_json", @@ -38,8 +38,8 @@ "status": "pass", "reason": "The evaluation protocol should appear before the deeper evidence ledger.", "scorecard_index": 37042, - "protocol_index": 41271, - "evidence_index": 43678 + "protocol_index": 41281, + "evidence_index": 43688 }, { "name": "evaluation_protocol_links_json", @@ -102,27 +102,27 @@ "json_files": [ { "path": "data/artifact_index.json", - "bytes": 20587, + "bytes": 22054, "top_level_type": "dict" }, { "path": "data/brand_assets.json", - "bytes": 3921, + "bytes": 3904, "top_level_type": "dict" }, { "path": "data/evaluation_protocol.json", - "bytes": 13575, + "bytes": 13586, "top_level_type": "dict" }, { "path": "data/evidence_contract.json", - "bytes": 10906, + "bytes": 10929, "top_level_type": "dict" }, { "path": "data/figure_index.json", - "bytes": 11493, + "bytes": 13449, "top_level_type": "dict" }, { @@ -132,7 +132,7 @@ }, { "path": "data/mirror_parity.json", - "bytes": 66049, + "bytes": 84322, "top_level_type": "dict" }, { @@ -147,12 +147,12 @@ }, { "path": "data/publication_audit.json", - "bytes": 6098, + "bytes": 6733, "top_level_type": "dict" }, { "path": "data/quality_gates.json", - "bytes": 5819, + "bytes": 6351, "top_level_type": "dict" }, { @@ -172,17 +172,17 @@ }, { "path": "data/reviewer_packet.json", - "bytes": 7386, + "bytes": 7401, "top_level_type": "dict" }, { "path": "data/reviewer_scorecard.json", - "bytes": 6107, + "bytes": 6114, "top_level_type": "dict" }, { "path": "data/scope_claims_audit.json", - "bytes": 19964, + "bytes": 20089, "top_level_type": "dict" }, { @@ -192,7 +192,7 @@ }, { "path": "data/summary_metrics.json", - "bytes": 25075, + "bytes": 25088, "top_level_type": "dict" }, { @@ -202,7 +202,7 @@ }, { "path": "data/website_integrity.json", - "bytes": 8528, + "bytes": 8813, "top_level_type": "dict" }, { diff --git a/docs/index.html b/docs/index.html index 3c3ef710bcc81b566307086eb5a1049378fbdbb6..761719ec34497ab5001b878ab2f985ea832ffdf9 100644 --- a/docs/index.html +++ b/docs/index.html @@ -1161,9 +1161,9 @@
data-gated

Qwen3-Omni pilot

-

The 32-episode LoRA path is prepared, but no model-quality claim is made until gated data access, held-out splits, training, and eval pass.

+

The 32-episode LoRA path is prepared, but no model-quality claim is made until gated data access, held-out splits, training, and evaluation pass.

- current claim smoke only + current claim readiness only target gate 32 episodes held-out eval pending
@@ -1209,7 +1209,7 @@

Evidence first, claims second.

-

A top-level project should make its proof boundary visible. This ledger separates verified single-episode artifacts from smoke-only Qwen3-Omni work and the pending 32-episode gate.

+

A top-level project should make its proof boundary visible. This ledger separates verified single-episode artifacts from readiness-only Qwen3-Omni work and the pending 32-episode gate.

@@ -1240,9 +1240,9 @@
- blocked by data access -

Qwen3-Omni is smoke-only until 32 episodes land

-

The H20/A100 pipeline exists, but current Qwen3-Omni artifacts use one episode and 128 train windows. No 32-episode metric is claimed.

+ data-gated +

Qwen3-Omni remains readiness-only

+

The current Qwen3-Omni artifacts use one episode and 128 train windows. No 32-episode metric is claimed.

evidence contract machine JSON @@ -1251,7 +1251,7 @@
verified

Scope claims are machine-checked

-

The audit confirms historical 32ep run/path strings stay confined to smoke-artifact provenance and are not presented as real 32-episode results.

+

The audit confirms historical 32ep run/path strings stay confined to readiness-artifact provenance and are not presented as real 32-episode results.

scope audit validator script @@ -1340,7 +1340,7 @@
01

Check the claim boundary

-

Start with the evidence contract, artifact index, scope audit, publication audit, and website integrity report. They separate verified single-episode artifacts from smoke-only Qwen3-Omni work.

+

Start with the evidence contract, artifact index, scope audit, publication audit, and website integrity report. They separate verified single-episode artifacts from readiness-only Qwen3-Omni work.

contract JSON @@ -1378,7 +1378,7 @@

The multi-episode Qwen3-Omni path is prepared, but no real 32-episode result is claimed until the data gate and held-out evaluation pass.

blocker - relay + access status review packet
@@ -1729,7 +1729,7 @@

Artifact guide

Human-readable map from proof boundary to data contract, task evidence, platform mirrors, and scale-up status.

ARTIFACT_GUIDE.md
-

Evidence contract

Defines verified, smoke-only, blocked, and out-of-scope claims.

EVIDENCE_CONTRACT.md
+

Evidence contract

Defines verified, readiness-only, blocked, and out-of-scope claims.

EVIDENCE_CONTRACT.md

Quality gates

One release checklist for automated validators and live post-publish checks.

quality_gates.json

Artifact index

Selective source-of-truth catalog with existence checks, sizes, and stable-file hashes.

artifact_index.json
@@ -1779,8 +1779,8 @@

The multi-episode Qwen3-Omni path is documented and scripted, but no full-pilot metric is claimed until the data gate and held-out evaluation pass.

-

A100 HF relay status

HF full-dataset access is pending; an A100 watcher is ready to download and transfer a 32-session pilot.

A100_HF_RELAY_STATUS.md
-

Qwen3-Omni readiness artifacts

Manifests, metadata, metrics, and progress logs from the current smoke/evidence run.

episode_manifest.json
+

Multi-episode access status

Public data-access boundary and selected 32-episode pilot plan, without private infrastructure details.

MULTI_EPISODE_ACCESS_STATUS.md
+

Qwen3-Omni readiness artifacts

Manifests, metadata, metrics, and progress logs from the current readiness run.

episode_manifest.json

32-episode data gate

The readiness gate remains the source of truth before any full pilot training claim.

DATA_BLOCKER_REPORT.md
@@ -1792,12 +1792,12 @@

Qwen3-Omni pilot is approval-ready.

-

The full Xperience-10M Hugging Face dataset is gated. While access is pending, the A100 relay has already selected a 32-episode pilot across 32 different session UUIDs and will continue automatically after approval.

+

The full Xperience-10M Hugging Face dataset is gated. While access is pending, the public plan has selected a 32-episode pilot across 32 different session UUIDs.

Selection

Stratified round-robin over 64 top-level sessions; 680 complete candidates scanned; 32 sessions selected.

-

Transfer

A100 downloads from Hugging Face, excludes visualization.rrd, validates files, then rsyncs to H20.

-

Boundary

The current LoRA artifact is a smoke/pilot checkpoint. A real 32-episode result requires the watcher to finish and held-out evaluation to run.

+

Transfer

Download raw episodes only from official gated sources, exclude visualization.rrd, validate files, then stage them for training.

+

Boundary

The current LoRA artifact is a readiness checkpoint. A real 32-episode result requires local gated data and held-out evaluation.

diff --git a/results/omni_exploration/modelscope_manifest.json b/results/omni_exploration/modelscope_manifest.json index ff4ea7486965817cf0d02776130ad3516bbeb0b6..6a66ab4c9128523a9c9961e4e4f74f5b90ee2e73 100644 --- a/results/omni_exploration/modelscope_manifest.json +++ b/results/omni_exploration/modelscope_manifest.json @@ -11,8 +11,8 @@ "episodes": [ { "episode_id": "xperience-10m-sample", - "path": "/home/cy/Ropedia/modelscope_data/xperience-10m-sample", - "annotation": "/home/cy/Ropedia/modelscope_data/xperience-10m-sample/annotation.hdf5", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample", + "annotation": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/annotation.hdf5", "files": [ { "name": "annotation.hdf5", diff --git a/results/omni_finetune/DATA_BLOCKER_REPORT.md b/results/omni_finetune/DATA_BLOCKER_REPORT.md new file mode 100644 index 0000000000000000000000000000000000000000..38b1fd5996c77039c06bddf47172f24b2cbc5bf7 --- /dev/null +++ b/results/omni_finetune/DATA_BLOCKER_REPORT.md @@ -0,0 +1,21 @@ +# Xperience-10M Fine-Tune Readiness + +Target episodes: 32 +Ready for 32-episode pilot: False +Selected source: none + +## Source counts +- local (degraded-valid): 1 / 1 +- modelscope (degraded-valid): 0 / 0 +- huggingface (degraded-valid): 0 / 0 + +## Blockers +- Not enough degraded-valid episodes for a 32-episode pilot. Need 32, local has 1. +- Current training host path remains one-episode proof-of-stack only. +- ModelScope probe unavailable or reported no matching episode files. +- Hugging Face probe unavailable or reported no matching episode files. + +## Interpretation +- Degraded-valid means: annotation.hdf5 and fisheye_cam0.mp4 both exist. +- Complete means all six MP4 views are present with annotation. +- A 32-episode pilot must not be claimed unless this script selects a source with 32+ degraded-valid episodes. diff --git a/results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md b/results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md new file mode 100644 index 0000000000000000000000000000000000000000..5923390e3b7174e97f419ee2834091981443fc00 --- /dev/null +++ b/results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md @@ -0,0 +1,39 @@ +# Multi-Episode Access Status + +Current blocker: access to the gated full `ropedia-ai/xperience-10m` dataset is +still pending approval from the dataset authors. + +This file records only public-facing readiness facts. It intentionally excludes +machine aliases, private paths, SSH hosts, token locations, and local server +details. + +## Selection Plan + +| Item | Value | +| --- | ---: | +| Dataset | `ropedia-ai/xperience-10m` | +| Target | 32 complete leaf episodes | +| Strategy | stratified round-robin across top-level session UUIDs | +| Candidate scan | first 64 top-level session UUIDs | +| Valid candidates | 680 | +| Selected sessions | 32 | +| Minimum episode size | 0.25 GB | +| Estimated bytes | 72,031,620,552 | +| Excluded file | `visualization.rrd` | + +## Boundary + +The current Qwen3-Omni artifacts are readiness artifacts from the locally +available sample data. They are not 32-episode held-out model-quality results. + +A real 32-episode pilot can be claimed only after: + +- at least 32 valid episodes are available locally, +- the manifest builder confirms complete held-out episode splits, +- training finishes with recorded metadata and progress logs, +- evaluation runs on held-out test episodes, +- predictions, metrics, confusion matrices, and a run report are committed. + +The source-of-truth blocker report remains: + +`results/omni_finetune/DATA_BLOCKER_REPORT.md` diff --git a/results/omni_finetune/RUN_REPORT.md b/results/omni_finetune/RUN_REPORT.md new file mode 100644 index 0000000000000000000000000000000000000000..62fdd21e5e38ba8f81cede2096be1c71a0133dc9 --- /dev/null +++ b/results/omni_finetune/RUN_REPORT.md @@ -0,0 +1,13 @@ +# Qwen3-Omni LoRA Evaluation + +- Base model: `/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct` +- Adapter: `checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora` +- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl` +- Eval split: `train` +- Samples: `128` +- Episodes: `1` +- Accuracy: `0.0000` +- Macro-F1: `0.0000` +- Unseen eval labels: `0` + +Artifacts include `metrics.json`, `predictions.csv`, `per_class_metrics.csv`, and `confusion_matrix.csv`. diff --git a/results/omni_finetune/RUN_REPORT_eval.md b/results/omni_finetune/RUN_REPORT_eval.md new file mode 100644 index 0000000000000000000000000000000000000000..62fdd21e5e38ba8f81cede2096be1c71a0133dc9 --- /dev/null +++ b/results/omni_finetune/RUN_REPORT_eval.md @@ -0,0 +1,13 @@ +# Qwen3-Omni LoRA Evaluation + +- Base model: `/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct` +- Adapter: `checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora` +- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl` +- Eval split: `train` +- Samples: `128` +- Episodes: `1` +- Accuracy: `0.0000` +- Macro-F1: `0.0000` +- Unseen eval labels: `0` + +Artifacts include `metrics.json`, `predictions.csv`, `per_class_metrics.csv`, and `confusion_matrix.csv`. diff --git a/results/omni_finetune/RUN_REPORT_lora.md b/results/omni_finetune/RUN_REPORT_lora.md new file mode 100644 index 0000000000000000000000000000000000000000..0fbbcadf95bab3949b07ec08c6b571d926750180 --- /dev/null +++ b/results/omni_finetune/RUN_REPORT_lora.md @@ -0,0 +1,11 @@ +# Qwen3-Omni LoRA Training + +- Base model: `/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct` +- Dataset: `results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl` +- Train samples: `128` +- Validation samples: `0` +- Processes: `8` +- Epochs: `1` +- Final train loss: `10.936364` + +Only LoRA parameters are trained; the base Qwen3-Omni weights remain frozen. diff --git a/results/omni_finetune/adapter_lora/README.md b/results/omni_finetune/adapter_lora/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7daea40193544e018122f7a2bb1a91aae8fd7e96 --- /dev/null +++ b/results/omni_finetune/adapter_lora/README.md @@ -0,0 +1,206 @@ +--- +base_model: '' +library_name: peft +tags: +- 'base_model:adapter:' +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.18.1 \ No newline at end of file diff --git a/results/omni_finetune/adapter_lora/adapter_config.json b/results/omni_finetune/adapter_lora/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..ef8ca4226f1b89d8b01f981e04775fb87c7e0543 --- /dev/null +++ b/results/omni_finetune/adapter_lora/adapter_config.json @@ -0,0 +1,49 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": { + "base_model_class": "Qwen3OmniMoeThinkerForConditionalGeneration", + "parent_library": "transformers.models.qwen3_omni_moe.modeling_qwen3_omni_moe" + }, + "base_model_name_or_path": "", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "peft_version": "0.18.1", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "up_proj", + "q_proj", + "v_proj", + "k_proj", + "down_proj", + "o_proj" + ], + "target_parameters": null, + "task_type": null, + "trainable_token_indices": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/results/omni_finetune/adapter_lora/chat_template.jinja b/results/omni_finetune/adapter_lora/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..1efcce1402ae5d1d61d4a4cf56accea5921251cf --- /dev/null +++ b/results/omni_finetune/adapter_lora/chat_template.jinja @@ -0,0 +1,122 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {{- "<|vision_start|><|image_pad|><|vision_end|>" }} + {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %} + {{- "<|audio_start|><|audio_pad|><|audio_end|>" }} + {%- elif content.type == 'video' or 'video' in content %} + {{- "<|vision_start|><|video_pad|><|vision_end|>" }} + {%- elif content.type == 'text' %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- endif %} + {{- '\n\n' }} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }} + {%- else %} + {%- for content in messages[0].content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {{- '<|im_start|>system\n' +"<|vision_start|><|image_pad|><|vision_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %} + {{- '<|im_start|>system\n' +"<|audio_start|><|audio_pad|><|audio_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'video' or 'video' in content %} + {{- '<|im_start|>system\n' +"<|vision_start|><|video_pad|><|vision_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'text' %} + {{- '<|im_start|>system\n' +content.text+ '<|im_end|>\n' }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} +{%- endfor %} +{%- for message in messages %} + {%- if message.content is string %} + {%- set content = message.content %} + {%- else %} + {%- set content = namespace(text="") %} + {%- for mcontent in message.content %} + {%- if mcontent.type == 'image' or 'image' in mcontent or 'image_url' in mcontent %} + {%- set content.text = content.text~"<|vision_start|><|image_pad|><|vision_end|>" %} + {%- elif mcontent.type == 'audio' or 'audio' in mcontent or 'audio_url' in mcontent %} + {%- set content.text = content.text~"<|audio_start|><|audio_pad|><|audio_end|>" %} + {%- elif mcontent.type == 'video' or 'video' in mcontent %} + {%- set content.text = content.text~"<|vision_start|><|video_pad|><|vision_end|>" %} + {%- elif mcontent.type == 'text' %} + {%- set content.text = content.text~mcontent.text %} + {%- endif %} + {%- endfor %} + {%- set content = content.text %} + {%- endif %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = "" %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- if loop.index0 > ns.last_query_index %} + {%- if loop.last or (not loop.last and reasoning_content) %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip("\n") + '\n\n\n' + content.lstrip('\n') }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and content) or (not loop.first) %}{{- '\n' }}{%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}{{- '<|im_start|>user' }}{%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}{{- '<|im_end|>\n' }}{%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %}{{- '\n\n\n\n' }}{%- endif %} +{%- endif %} \ No newline at end of file diff --git a/results/omni_finetune/adapter_lora/processor_config.json b/results/omni_finetune/adapter_lora/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..f5e20e1e0b4ab8f002e2475fc2d49e8910d5c627 --- /dev/null +++ b/results/omni_finetune/adapter_lora/processor_config.json @@ -0,0 +1,114 @@ +{ + "feature_extractor": { + "chunk_length": 30, + "dither": 0.0, + "feature_extractor_type": "WhisperFeatureExtractor", + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_pixels": 12845056, + "merge_size": 2, + "min_pixels": 3136, + "n_fft": 400, + "n_samples": 480000, + "nb_max_frames": 3000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "return_attention_mask": true, + "sampling_rate": 16000, + "temporal_patch_size": 2 + }, + "image_processor": { + "data_format": "channels_first", + "dither": 0.0, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "n_fft": 400, + "n_samples": 4800000, + "nb_max_frames": 30000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_attention_mask": true, + "sampling_rate": 16000, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3OmniMoeProcessor", + "video_processor": { + "data_format": "channels_first", + "default_to_square": true, + "dither": 0.0, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": false, + "feature_extractor_type": "WhisperFeatureExtractor", + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "n_fft": 400, + "n_samples": 4800000, + "nb_max_frames": 30000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_attention_mask": true, + "return_metadata": false, + "sampling_rate": 16000, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen2VLVideoProcessor" + } +} diff --git a/results/omni_finetune/adapter_lora/tokenizer.json b/results/omni_finetune/adapter_lora/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f7ab87da8ac9fbb6ebdd03fb50597a92933ce3c6 --- /dev/null +++ b/results/omni_finetune/adapter_lora/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d41145c408f3062e824f965be1c29854cd809e111cdf8170aa8b0bcd5d1fab +size 11424262 diff --git a/results/omni_finetune/adapter_lora/tokenizer_config.json b/results/omni_finetune/adapter_lora/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..22039d0853e5588d2b04702b4dbde3ff8546edfc --- /dev/null +++ b/results/omni_finetune/adapter_lora/tokenizer_config.json @@ -0,0 +1,52 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>", + "<|audio_start|>", + "<|audio_end|>", + "", + "", + "", + "<|audio_pad|>" + ], + "image_token": "<|image_pad|>", + "is_local": true, + "model_max_length": 131072, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3OmniMoeProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/results/omni_finetune/adapter_lora/training_metadata.json b/results/omni_finetune/adapter_lora/training_metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..50814ee383d9cf91fe31fbdd5e76a2cdacff9215 --- /dev/null +++ b/results/omni_finetune/adapter_lora/training_metadata.json @@ -0,0 +1,32 @@ +{ + "run_id": "xperience10m_qwen3_omni_32ep_lora", + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "num_processes": 8, + "num_train_samples": 128, + "num_val_samples": 0, + "history": [ + { + "epoch": 1, + "train_loss": 10.936363816261292, + "val_loss": null, + "global_step": 16 + } + ], + "lora": { + "r": 16, + "alpha": 32, + "dropout": 0.05, + "target_modules": [ + "q_proj", + "k_proj", + "v_proj", + "o_proj", + "gate_proj", + "up_proj", + "down_proj" + ] + }, + "use_audio_in_video": false +} \ No newline at end of file diff --git a/results/omni_finetune/config.yaml b/results/omni_finetune/config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..22f96748a9d9ec2cc8bd8f73d5bad1f6851ef5d3 --- /dev/null +++ b/results/omni_finetune/config.yaml @@ -0,0 +1,10 @@ +run_id: xperience10m_qwen3_omni_32ep_lora +stage: qwen_lora_text_video_audio +model_id: /path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct +dataset_jsonl: results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl +checkpoint_dir: /path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora +num_processes: 8 +epochs: 1 +learning_rate: 0.0001 +lora_r: 16 +lora_alpha: 32 diff --git a/results/omni_finetune/confusion_matrix_eval.csv b/results/omni_finetune/confusion_matrix_eval.csv new file mode 100644 index 0000000000000000000000000000000000000000..756ef35187a9682805dd8c748605e41eff661209 --- /dev/null +++ b/results/omni_finetune/confusion_matrix_eval.csv @@ -0,0 +1,20 @@ +true\pred,Close bottle cap,Grasp coffee scoop,Grasp gooseneck kettle,Hold coffee carafe,Hold gooseneck kettle,Lift gooseneck kettle,Move kettle,Move kettle away,Pick up kettle,Pick up white bottle,Place item on table,Place kettle on table,Position kettle to pour,Pour coffee,Pour liquid from white bottle,Pour milk into coffee,Transfer coffee to dripper,Wait/Prepare for pouring,unknown +Close bottle cap,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,3,5 +Grasp coffee scoop,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,6 +Grasp gooseneck kettle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,5 +Hold coffee carafe,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,7 +Hold gooseneck kettle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8 +Lift gooseneck kettle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,6 +Move kettle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,7 +Move kettle away,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8 +Pick up kettle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,6,2 +Pick up white bottle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,2 +Place item on table,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,4 +Place kettle on table,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,3,3 +Position kettle to pour,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,7,1 +Pour coffee,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,7 +Pour liquid from white bottle,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,2 +Pour milk into coffee,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1 +Transfer coffee to dripper,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,3,5 +Wait/Prepare for pouring,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8 +unknown,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 diff --git a/results/omni_finetune/dataset.jsonl b/results/omni_finetune/dataset.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..a4c28aed87287017bc931600e88ab9bccc4f0045 --- /dev/null +++ b/results/omni_finetune/dataset.jsonl @@ -0,0 +1,128 @@ +{"id": "xperience-10m-sample:qa:0", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 0, "end_frame": 19, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 0, "context_end_frame": 119, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00000_ctx0_119_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00000_ctx0_119_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 0, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 0, "end_frame": 19}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00000_ctx0_119_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00000_ctx0_119_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 0-19\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 19, \"start_frame\": 0}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:1", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 20, "end_frame": 39, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 0, "context_end_frame": 119, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00001_ctx0_119_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00001_ctx0_119_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 1, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 20, "end_frame": 39}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00001_ctx0_119_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00001_ctx0_119_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 20-39\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 39, \"start_frame\": 20}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:2", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 40, "end_frame": 59, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 0, "context_end_frame": 119, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00002_ctx0_119_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00002_ctx0_119_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 2, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 40, "end_frame": 59}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00002_ctx0_119_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00002_ctx0_119_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 40-59\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 59, \"start_frame\": 40}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:3", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 60, "end_frame": 79, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 9, "context_end_frame": 128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00003_ctx9_128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00003_ctx9_128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 3, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 60, "end_frame": 79}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00003_ctx9_128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00003_ctx9_128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 60-79\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 79, \"start_frame\": 60}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:4", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 80, "end_frame": 99, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 29, "context_end_frame": 148, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00004_ctx29_148_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00004_ctx29_148_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 4, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 80, "end_frame": 99}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00004_ctx29_148_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00004_ctx29_148_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 80-99\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 99, \"start_frame\": 80}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:5", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 100, "end_frame": 119, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 49, "context_end_frame": 168, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00005_ctx49_168_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00005_ctx49_168_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 5, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 100, "end_frame": 119}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00005_ctx49_168_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00005_ctx49_168_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 100-119\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 119, \"start_frame\": 100}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\", \"coffee dripper\", \"scale\", \"bottle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:6", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 120, "end_frame": 139, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 69, "context_end_frame": 188, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00006_ctx69_188_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00006_ctx69_188_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 6, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 120, "end_frame": 139}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00006_ctx69_188_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00006_ctx69_188_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 120-139\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 139, \"start_frame\": 120}, \"next_action\": \"Pick up kettle\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:7", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 140, "end_frame": 159, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 89, "context_end_frame": 208, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00007_ctx89_208_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00007_ctx89_208_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 7, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 140, "end_frame": 159}}, "label": "Pick up kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00007_ctx89_208_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00007_ctx89_208_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 140-159\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 159, \"start_frame\": 140}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:8", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 160, "end_frame": 179, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 109, "context_end_frame": 228, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00008_ctx109_228_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00008_ctx109_228_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 8, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 160, "end_frame": 179}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00008_ctx109_228_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00008_ctx109_228_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 160-179\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 179, \"start_frame\": 160}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:9", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 180, "end_frame": 199, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 129, "context_end_frame": 248, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00009_ctx129_248_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00009_ctx129_248_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 9, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 180, "end_frame": 199}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00009_ctx129_248_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00009_ctx129_248_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 180-199\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 199, \"start_frame\": 180}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:10", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 200, "end_frame": 219, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 149, "context_end_frame": 268, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00010_ctx149_268_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00010_ctx149_268_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 10, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 200, "end_frame": 219}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00010_ctx149_268_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00010_ctx149_268_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 200-219\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 219, \"start_frame\": 200}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:11", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 220, "end_frame": 239, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 169, "context_end_frame": 288, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00011_ctx169_288_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00011_ctx169_288_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 11, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 220, "end_frame": 239}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00011_ctx169_288_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00011_ctx169_288_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 220-239\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 239, \"start_frame\": 220}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:12", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 240, "end_frame": 259, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 189, "context_end_frame": 308, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00012_ctx189_308_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00012_ctx189_308_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 12, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 240, "end_frame": 259}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00012_ctx189_308_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00012_ctx189_308_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 240-259\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 259, \"start_frame\": 240}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:13", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 260, "end_frame": 279, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 209, "context_end_frame": 328, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00013_ctx209_328_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00013_ctx209_328_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 13, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 260, "end_frame": 279}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00013_ctx209_328_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00013_ctx209_328_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 260-279\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 279, \"start_frame\": 260}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:14", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 280, "end_frame": 299, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 229, "context_end_frame": 348, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00014_ctx229_348_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00014_ctx229_348_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 14, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 280, "end_frame": 299}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00014_ctx229_348_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00014_ctx229_348_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 280-299\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 299, \"start_frame\": 280}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:15", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 300, "end_frame": 319, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 249, "context_end_frame": 368, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00015_ctx249_368_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00015_ctx249_368_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 15, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 300, "end_frame": 319}}, "label": "Position kettle to pour", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00015_ctx249_368_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00015_ctx249_368_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 300-319\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Position kettle to pour\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 319, \"start_frame\": 300}, \"next_action\": \"Position kettle to pour\", \"objects\": [\"kettle\"], \"subtask\": \"Pick up and position kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:40", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 800, "end_frame": 819, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 749, "context_end_frame": 868, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00040_ctx749_868_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00040_ctx749_868_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 40, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 800, "end_frame": 819}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00040_ctx749_868_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00040_ctx749_868_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 800-819\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 819, \"start_frame\": 800}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:41", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 820, "end_frame": 839, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 769, "context_end_frame": 888, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00041_ctx769_888_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00041_ctx769_888_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 41, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 820, "end_frame": 839}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00041_ctx769_888_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00041_ctx769_888_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 820-839\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 839, \"start_frame\": 820}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:42", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 840, "end_frame": 859, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 789, "context_end_frame": 908, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00042_ctx789_908_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00042_ctx789_908_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 42, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 840, "end_frame": 859}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00042_ctx789_908_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00042_ctx789_908_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 840-859\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 859, \"start_frame\": 840}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:43", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 860, "end_frame": 879, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 809, "context_end_frame": 928, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00043_ctx809_928_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00043_ctx809_928_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 43, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 860, "end_frame": 879}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00043_ctx809_928_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00043_ctx809_928_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 860-879\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 879, \"start_frame\": 860}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:44", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 880, "end_frame": 899, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 829, "context_end_frame": 948, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00044_ctx829_948_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00044_ctx829_948_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 44, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 880, "end_frame": 899}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00044_ctx829_948_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00044_ctx829_948_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 880-899\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 899, \"start_frame\": 880}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:45", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 900, "end_frame": 919, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 849, "context_end_frame": 968, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00045_ctx849_968_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00045_ctx849_968_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 45, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 900, "end_frame": 919}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00045_ctx849_968_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00045_ctx849_968_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 900-919\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 919, \"start_frame\": 900}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:46", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 920, "end_frame": 939, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 869, "context_end_frame": 988, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00046_ctx869_988_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00046_ctx869_988_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 46, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 920, "end_frame": 939}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00046_ctx869_988_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00046_ctx869_988_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 920-939\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 939, \"start_frame\": 920}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:47", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 940, "end_frame": 959, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 889, "context_end_frame": 1008, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00047_ctx889_1008_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00047_ctx889_1008_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 47, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 940, "end_frame": 959}}, "label": "Move kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00047_ctx889_1008_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00047_ctx889_1008_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 940-959\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 959, \"start_frame\": 940}, \"next_action\": \"Move kettle\", \"objects\": [\"kettle\", \"dripper\", \"scale\", \"coffee filter\", \"table\"], \"subtask\": \"Move kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:50", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1000, "end_frame": 1019, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 949, "context_end_frame": 1068, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00050_ctx949_1068_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00050_ctx949_1068_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 50, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1000, "end_frame": 1019}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00050_ctx949_1068_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00050_ctx949_1068_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1000-1019\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1019, \"start_frame\": 1000}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:51", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1020, "end_frame": 1039, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 969, "context_end_frame": 1088, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00051_ctx969_1088_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00051_ctx969_1088_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 51, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1020, "end_frame": 1039}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00051_ctx969_1088_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00051_ctx969_1088_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1020-1039\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1039, \"start_frame\": 1020}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:52", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1040, "end_frame": 1059, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 989, "context_end_frame": 1108, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00052_ctx989_1108_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00052_ctx989_1108_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 52, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1040, "end_frame": 1059}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00052_ctx989_1108_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00052_ctx989_1108_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1040-1059\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1059, \"start_frame\": 1040}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:53", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1060, "end_frame": 1079, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1009, "context_end_frame": 1128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00053_ctx1009_1128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00053_ctx1009_1128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 53, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1060, "end_frame": 1079}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00053_ctx1009_1128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00053_ctx1009_1128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1060-1079\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1079, \"start_frame\": 1060}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:54", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1080, "end_frame": 1099, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1029, "context_end_frame": 1148, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00054_ctx1029_1148_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00054_ctx1029_1148_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 54, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1080, "end_frame": 1099}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00054_ctx1029_1148_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00054_ctx1029_1148_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1080-1099\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1099, \"start_frame\": 1080}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:55", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1100, "end_frame": 1119, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1049, "context_end_frame": 1168, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00055_ctx1049_1168_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00055_ctx1049_1168_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 55, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1100, "end_frame": 1119}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00055_ctx1049_1168_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00055_ctx1049_1168_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1100-1119\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1119, \"start_frame\": 1100}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:56", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1120, "end_frame": 1139, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1069, "context_end_frame": 1188, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00056_ctx1069_1188_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00056_ctx1069_1188_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 56, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1120, "end_frame": 1139}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00056_ctx1069_1188_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00056_ctx1069_1188_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1120-1139\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1139, \"start_frame\": 1120}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:57", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1140, "end_frame": 1159, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1089, "context_end_frame": 1208, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00057_ctx1089_1208_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00057_ctx1089_1208_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 57, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1140, "end_frame": 1159}}, "label": "Hold coffee carafe", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00057_ctx1089_1208_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00057_ctx1089_1208_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1140-1159\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold coffee carafe\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1159, \"start_frame\": 1140}, \"next_action\": \"Hold coffee carafe\", \"objects\": [\"glass carafe\", \"coffee scale\", \"coffee dripper\", \"white mug\", \"bottle\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:62", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1240, "end_frame": 1259, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1189, "context_end_frame": 1308, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00062_ctx1189_1308_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00062_ctx1189_1308_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 62, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1240, "end_frame": 1259}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00062_ctx1189_1308_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00062_ctx1189_1308_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1240-1259\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1259, \"start_frame\": 1240}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:63", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1260, "end_frame": 1279, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1209, "context_end_frame": 1328, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00063_ctx1209_1328_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00063_ctx1209_1328_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 63, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1260, "end_frame": 1279}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00063_ctx1209_1328_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00063_ctx1209_1328_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1260-1279\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1279, \"start_frame\": 1260}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:64", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1280, "end_frame": 1299, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1229, "context_end_frame": 1348, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00064_ctx1229_1348_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00064_ctx1229_1348_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 64, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1280, "end_frame": 1299}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00064_ctx1229_1348_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00064_ctx1229_1348_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1280-1299\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1299, \"start_frame\": 1280}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:65", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1300, "end_frame": 1319, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1249, "context_end_frame": 1368, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00065_ctx1249_1368_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00065_ctx1249_1368_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 65, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1300, "end_frame": 1319}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00065_ctx1249_1368_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00065_ctx1249_1368_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1300-1319\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1319, \"start_frame\": 1300}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:66", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1320, "end_frame": 1339, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1269, "context_end_frame": 1388, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00066_ctx1269_1388_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00066_ctx1269_1388_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 66, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1320, "end_frame": 1339}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00066_ctx1269_1388_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00066_ctx1269_1388_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1320-1339\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1339, \"start_frame\": 1320}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:67", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1340, "end_frame": 1359, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1289, "context_end_frame": 1408, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00067_ctx1289_1408_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00067_ctx1289_1408_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 67, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1340, "end_frame": 1359}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00067_ctx1289_1408_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00067_ctx1289_1408_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1340-1359\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1359, \"start_frame\": 1340}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:68", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1360, "end_frame": 1379, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1309, "context_end_frame": 1428, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00068_ctx1309_1428_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00068_ctx1309_1428_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 68, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1360, "end_frame": 1379}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00068_ctx1309_1428_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00068_ctx1309_1428_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1360-1379\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1379, \"start_frame\": 1360}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:69", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1380, "end_frame": 1399, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1329, "context_end_frame": 1448, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00069_ctx1329_1448_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00069_ctx1329_1448_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 69, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1380, "end_frame": 1399}}, "label": "Grasp coffee scoop", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00069_ctx1329_1448_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00069_ctx1329_1448_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1380-1399\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp coffee scoop\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1399, \"start_frame\": 1380}, \"next_action\": \"Grasp coffee scoop\", \"objects\": [\"glass carafe\", \"wooden scoop\"], \"subtask\": \"Prepare coffee equipment and scoop grounds\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:78", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1560, "end_frame": 1579, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1509, "context_end_frame": 1628, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00078_ctx1509_1628_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00078_ctx1509_1628_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 78, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee jar", "coffee scoop"], "contact": "yes", "transition": "yes", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1560, "end_frame": 1579}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00078_ctx1509_1628_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00078_ctx1509_1628_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1560-1579\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1579, \"start_frame\": 1560}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee jar\", \"coffee scoop\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"yes\"}"}]} +{"id": "xperience-10m-sample:qa:79", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1580, "end_frame": 1599, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1529, "context_end_frame": 1648, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00079_ctx1529_1648_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00079_ctx1529_1648_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 79, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee jar", "coffee scoop"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1580, "end_frame": 1599}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00079_ctx1529_1648_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00079_ctx1529_1648_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1580-1599\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1599, \"start_frame\": 1580}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee jar\", \"coffee scoop\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:80", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1600, "end_frame": 1619, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1549, "context_end_frame": 1668, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00080_ctx1549_1668_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00080_ctx1549_1668_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 80, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1600, "end_frame": 1619}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00080_ctx1549_1668_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00080_ctx1549_1668_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1600-1619\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1619, \"start_frame\": 1600}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:81", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1620, "end_frame": 1639, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1569, "context_end_frame": 1688, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00081_ctx1569_1688_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00081_ctx1569_1688_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 81, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1620, "end_frame": 1639}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00081_ctx1569_1688_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00081_ctx1569_1688_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1620-1639\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1639, \"start_frame\": 1620}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:82", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1640, "end_frame": 1659, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1589, "context_end_frame": 1708, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00082_ctx1589_1708_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00082_ctx1589_1708_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 82, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1640, "end_frame": 1659}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00082_ctx1589_1708_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00082_ctx1589_1708_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1640-1659\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1659, \"start_frame\": 1640}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:83", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1660, "end_frame": 1679, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1609, "context_end_frame": 1728, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00083_ctx1609_1728_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00083_ctx1609_1728_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 83, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1660, "end_frame": 1679}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00083_ctx1609_1728_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00083_ctx1609_1728_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1660-1679\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1679, \"start_frame\": 1660}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:84", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1680, "end_frame": 1699, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1629, "context_end_frame": 1748, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00084_ctx1629_1748_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00084_ctx1629_1748_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 84, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1680, "end_frame": 1699}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00084_ctx1629_1748_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00084_ctx1629_1748_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1680-1699\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1699, \"start_frame\": 1680}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:85", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 1700, "end_frame": 1719, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1649, "context_end_frame": 1768, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00085_ctx1649_1768_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00085_ctx1649_1768_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 85, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1700, "end_frame": 1719}}, "label": "Transfer coffee to dripper", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00085_ctx1649_1768_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00085_ctx1649_1768_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 1700-1719\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Transfer coffee to dripper\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 1719, \"start_frame\": 1700}, \"next_action\": \"Transfer coffee to dripper\", \"objects\": [\"coffee scoop\", \"dripper\"], \"subtask\": \"Transfer coffee grounds to dripper\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:100", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2000, "end_frame": 2019, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1949, "context_end_frame": 2068, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00100_ctx1949_2068_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00100_ctx1949_2068_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 100, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2000, "end_frame": 2019}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00100_ctx1949_2068_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00100_ctx1949_2068_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2000-2019\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2019, \"start_frame\": 2000}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:101", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2020, "end_frame": 2039, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1969, "context_end_frame": 2088, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00101_ctx1969_2088_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00101_ctx1969_2088_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 101, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2020, "end_frame": 2039}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00101_ctx1969_2088_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00101_ctx1969_2088_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2020-2039\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2039, \"start_frame\": 2020}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:102", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2040, "end_frame": 2059, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 1989, "context_end_frame": 2108, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00102_ctx1989_2108_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00102_ctx1989_2108_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 102, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2040, "end_frame": 2059}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00102_ctx1989_2108_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00102_ctx1989_2108_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2040-2059\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2059, \"start_frame\": 2040}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:103", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2060, "end_frame": 2079, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2009, "context_end_frame": 2128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00103_ctx2009_2128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00103_ctx2009_2128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 103, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2060, "end_frame": 2079}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00103_ctx2009_2128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00103_ctx2009_2128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2060-2079\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2079, \"start_frame\": 2060}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:104", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2080, "end_frame": 2099, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2029, "context_end_frame": 2148, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00104_ctx2029_2148_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00104_ctx2029_2148_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 104, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2080, "end_frame": 2099}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00104_ctx2029_2148_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00104_ctx2029_2148_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2080-2099\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2099, \"start_frame\": 2080}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:105", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2100, "end_frame": 2119, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2049, "context_end_frame": 2168, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00105_ctx2049_2168_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00105_ctx2049_2168_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 105, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2100, "end_frame": 2119}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00105_ctx2049_2168_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00105_ctx2049_2168_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2100-2119\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2119, \"start_frame\": 2100}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:106", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2120, "end_frame": 2139, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2069, "context_end_frame": 2188, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00106_ctx2069_2188_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00106_ctx2069_2188_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 106, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2120, "end_frame": 2139}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00106_ctx2069_2188_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00106_ctx2069_2188_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2120-2139\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2139, \"start_frame\": 2120}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:107", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2140, "end_frame": 2159, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2089, "context_end_frame": 2208, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00107_ctx2089_2208_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00107_ctx2089_2208_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 107, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2140, "end_frame": 2159}}, "label": "Hold gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00107_ctx2089_2208_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00107_ctx2089_2208_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2140-2159\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Hold gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2159, \"start_frame\": 2140}, \"next_action\": \"Hold gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"water bottle\", \"coffee mug\"], \"subtask\": \"Handle gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:140", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2800, "end_frame": 2819, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2749, "context_end_frame": 2868, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00140_ctx2749_2868_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00140_ctx2749_2868_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 140, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2800, "end_frame": 2819}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00140_ctx2749_2868_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00140_ctx2749_2868_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2800-2819\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2819, \"start_frame\": 2800}, \"next_action\": \"Grasp gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:141", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2820, "end_frame": 2839, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2769, "context_end_frame": 2888, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00141_ctx2769_2888_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00141_ctx2769_2888_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 141, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2820, "end_frame": 2839}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00141_ctx2769_2888_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00141_ctx2769_2888_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2820-2839\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2839, \"start_frame\": 2820}, \"next_action\": \"Grasp gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:142", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2840, "end_frame": 2859, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2789, "context_end_frame": 2908, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00142_ctx2789_2908_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00142_ctx2789_2908_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 142, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2840, "end_frame": 2859}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00142_ctx2789_2908_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00142_ctx2789_2908_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2840-2859\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2859, \"start_frame\": 2840}, \"next_action\": \"Grasp gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:143", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2860, "end_frame": 2879, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2809, "context_end_frame": 2928, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00143_ctx2809_2928_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00143_ctx2809_2928_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 143, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2860, "end_frame": 2879}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00143_ctx2809_2928_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00143_ctx2809_2928_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2860-2879\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2879, \"start_frame\": 2860}, \"next_action\": \"Grasp gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:144", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2880, "end_frame": 2899, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2829, "context_end_frame": 2948, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00144_ctx2829_2948_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00144_ctx2829_2948_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 144, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2880, "end_frame": 2899}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00144_ctx2829_2948_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00144_ctx2829_2948_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2880-2899\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2899, \"start_frame\": 2880}, \"next_action\": \"Grasp gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:145", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2900, "end_frame": 2919, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2849, "context_end_frame": 2968, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00145_ctx2849_2968_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00145_ctx2849_2968_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 145, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2900, "end_frame": 2919}}, "label": "Grasp gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00145_ctx2849_2968_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00145_ctx2849_2968_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2900-2919\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Grasp gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2919, \"start_frame\": 2900}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:146", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2920, "end_frame": 2939, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2869, "context_end_frame": 2988, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00146_ctx2869_2988_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00146_ctx2869_2988_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 146, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2920, "end_frame": 2939}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00146_ctx2869_2988_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00146_ctx2869_2988_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2920-2939\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2939, \"start_frame\": 2920}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:147", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2940, "end_frame": 2959, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2889, "context_end_frame": 3008, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00147_ctx2889_3008_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00147_ctx2889_3008_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 147, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2940, "end_frame": 2959}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00147_ctx2889_3008_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00147_ctx2889_3008_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2940-2959\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2959, \"start_frame\": 2940}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"water bottle\", \"white cup\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:148", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2960, "end_frame": 2979, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2909, "context_end_frame": 3028, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00148_ctx2909_3028_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00148_ctx2909_3028_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 148, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2960, "end_frame": 2979}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00148_ctx2909_3028_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00148_ctx2909_3028_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2960-2979\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2979, \"start_frame\": 2960}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:149", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 2980, "end_frame": 2999, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2929, "context_end_frame": 3048, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00149_ctx2929_3048_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00149_ctx2929_3048_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 149, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2980, "end_frame": 2999}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00149_ctx2929_3048_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00149_ctx2929_3048_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 2980-2999\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 2999, \"start_frame\": 2980}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:150", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3000, "end_frame": 3019, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2949, "context_end_frame": 3068, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00150_ctx2949_3068_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00150_ctx2949_3068_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 150, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3000, "end_frame": 3019}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00150_ctx2949_3068_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00150_ctx2949_3068_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3000-3019\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3019, \"start_frame\": 3000}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"white mug\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:151", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3020, "end_frame": 3039, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2969, "context_end_frame": 3088, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00151_ctx2969_3088_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00151_ctx2969_3088_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 151, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3020, "end_frame": 3039}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00151_ctx2969_3088_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00151_ctx2969_3088_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3020-3039\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3039, \"start_frame\": 3020}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"white mug\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:152", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3040, "end_frame": 3059, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 2989, "context_end_frame": 3108, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00152_ctx2989_3108_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00152_ctx2989_3108_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 152, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3040, "end_frame": 3059}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00152_ctx2989_3108_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00152_ctx2989_3108_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3040-3059\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3059, \"start_frame\": 3040}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"white mug\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:153", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3060, "end_frame": 3079, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3009, "context_end_frame": 3128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00153_ctx3009_3128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00153_ctx3009_3128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 153, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3060, "end_frame": 3079}}, "label": "Lift gooseneck kettle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00153_ctx3009_3128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00153_ctx3009_3128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3060-3079\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Lift gooseneck kettle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3079, \"start_frame\": 3060}, \"next_action\": \"Lift gooseneck kettle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"digital scale\", \"white mug\"], \"subtask\": \"Lift gooseneck kettle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:170", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3400, "end_frame": 3419, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3349, "context_end_frame": 3468, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00170_ctx3349_3468_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00170_ctx3349_3468_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 170, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3400, "end_frame": 3419}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00170_ctx3349_3468_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00170_ctx3349_3468_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3400-3419\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3419, \"start_frame\": 3400}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:171", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3420, "end_frame": 3439, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3369, "context_end_frame": 3488, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00171_ctx3369_3488_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00171_ctx3369_3488_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 171, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3420, "end_frame": 3439}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00171_ctx3369_3488_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00171_ctx3369_3488_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3420-3439\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3439, \"start_frame\": 3420}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:172", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3440, "end_frame": 3459, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3389, "context_end_frame": 3508, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00172_ctx3389_3508_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00172_ctx3389_3508_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 172, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3440, "end_frame": 3459}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00172_ctx3389_3508_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00172_ctx3389_3508_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3440-3459\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3459, \"start_frame\": 3440}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:173", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3460, "end_frame": 3479, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3409, "context_end_frame": 3528, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00173_ctx3409_3528_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00173_ctx3409_3528_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 173, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3460, "end_frame": 3479}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00173_ctx3409_3528_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00173_ctx3409_3528_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3460-3479\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3479, \"start_frame\": 3460}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:174", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3480, "end_frame": 3499, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3429, "context_end_frame": 3548, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00174_ctx3429_3548_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00174_ctx3429_3548_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 174, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3480, "end_frame": 3499}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00174_ctx3429_3548_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00174_ctx3429_3548_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3480-3499\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3499, \"start_frame\": 3480}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:175", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3500, "end_frame": 3519, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3449, "context_end_frame": 3568, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00175_ctx3449_3568_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00175_ctx3449_3568_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 175, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3500, "end_frame": 3519}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00175_ctx3449_3568_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00175_ctx3449_3568_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3500-3519\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3519, \"start_frame\": 3500}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:176", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3520, "end_frame": 3539, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3469, "context_end_frame": 3588, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00176_ctx3469_3588_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00176_ctx3469_3588_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 176, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3520, "end_frame": 3539}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00176_ctx3469_3588_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00176_ctx3469_3588_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3520-3539\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3539, \"start_frame\": 3520}, \"next_action\": \"Move kettle away\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:177", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3540, "end_frame": 3559, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3489, "context_end_frame": 3608, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00177_ctx3489_3608_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00177_ctx3489_3608_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 177, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3540, "end_frame": 3559}}, "label": "Move kettle away", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00177_ctx3489_3608_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00177_ctx3489_3608_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3540-3559\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Move kettle away\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3559, \"start_frame\": 3540}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:178", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3560, "end_frame": 3579, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3509, "context_end_frame": 3628, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00178_ctx3509_3628_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00178_ctx3509_3628_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 178, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3560, "end_frame": 3579}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00178_ctx3509_3628_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00178_ctx3509_3628_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3560-3579\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3579, \"start_frame\": 3560}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:179", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3580, "end_frame": 3599, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3529, "context_end_frame": 3648, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00179_ctx3529_3648_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00179_ctx3529_3648_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 179, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3580, "end_frame": 3599}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00179_ctx3529_3648_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00179_ctx3529_3648_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3580-3599\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3599, \"start_frame\": 3580}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:180", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3600, "end_frame": 3619, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3549, "context_end_frame": 3668, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00180_ctx3549_3668_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00180_ctx3549_3668_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 180, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3600, "end_frame": 3619}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00180_ctx3549_3668_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00180_ctx3549_3668_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3600-3619\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3619, \"start_frame\": 3600}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:181", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3620, "end_frame": 3639, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3569, "context_end_frame": 3688, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00181_ctx3569_3688_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00181_ctx3569_3688_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 181, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3620, "end_frame": 3639}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00181_ctx3569_3688_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00181_ctx3569_3688_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3620-3639\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3639, \"start_frame\": 3620}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:182", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3640, "end_frame": 3659, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3589, "context_end_frame": 3708, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00182_ctx3589_3708_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00182_ctx3589_3708_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 182, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3640, "end_frame": 3659}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00182_ctx3589_3708_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00182_ctx3589_3708_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3640-3659\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3659, \"start_frame\": 3640}, \"next_action\": \"Place kettle on table\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:183", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3660, "end_frame": 3679, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3609, "context_end_frame": 3728, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00183_ctx3609_3728_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00183_ctx3609_3728_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 183, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3660, "end_frame": 3679}}, "label": "Place kettle on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00183_ctx3609_3728_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00183_ctx3609_3728_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3660-3679\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place kettle on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3679, \"start_frame\": 3660}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:184", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3680, "end_frame": 3699, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3629, "context_end_frame": 3748, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00184_ctx3629_3748_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00184_ctx3629_3748_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 184, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3680, "end_frame": 3699}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00184_ctx3629_3748_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00184_ctx3629_3748_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3680-3699\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3699, \"start_frame\": 3680}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:185", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3700, "end_frame": 3719, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3649, "context_end_frame": 3768, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00185_ctx3649_3768_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00185_ctx3649_3768_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 185, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3700, "end_frame": 3719}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00185_ctx3649_3768_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00185_ctx3649_3768_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3700-3719\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3719, \"start_frame\": 3700}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:186", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3720, "end_frame": 3739, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3669, "context_end_frame": 3788, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00186_ctx3669_3788_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00186_ctx3669_3788_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 186, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3720, "end_frame": 3739}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00186_ctx3669_3788_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00186_ctx3669_3788_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3720-3739\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3739, \"start_frame\": 3720}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:187", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3740, "end_frame": 3759, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3689, "context_end_frame": 3808, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00187_ctx3689_3808_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00187_ctx3689_3808_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 187, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3740, "end_frame": 3759}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00187_ctx3689_3808_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00187_ctx3689_3808_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3740-3759\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3759, \"start_frame\": 3740}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:188", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3760, "end_frame": 3779, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3709, "context_end_frame": 3828, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00188_ctx3709_3828_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00188_ctx3709_3828_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 188, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3760, "end_frame": 3779}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00188_ctx3709_3828_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00188_ctx3709_3828_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3760-3779\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3779, \"start_frame\": 3760}, \"next_action\": \"Pick up white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:189", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3780, "end_frame": 3799, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3729, "context_end_frame": 3848, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00189_ctx3729_3848_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00189_ctx3729_3848_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 189, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3780, "end_frame": 3799}}, "label": "Pick up white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00189_ctx3729_3848_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00189_ctx3729_3848_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3780-3799\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pick up white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3799, \"start_frame\": 3780}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"gooseneck kettle\", \"coffee dripper\", \"scale\", \"white bottle\"], \"subtask\": \"Set down kettle and retrieve white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:190", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3800, "end_frame": 3819, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3749, "context_end_frame": 3868, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00190_ctx3749_3868_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00190_ctx3749_3868_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 190, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3800, "end_frame": 3819}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00190_ctx3749_3868_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00190_ctx3749_3868_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3800-3819\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3819, \"start_frame\": 3800}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:191", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3820, "end_frame": 3839, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3769, "context_end_frame": 3888, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00191_ctx3769_3888_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00191_ctx3769_3888_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 191, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3820, "end_frame": 3839}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00191_ctx3769_3888_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00191_ctx3769_3888_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3820-3839\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3839, \"start_frame\": 3820}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:192", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3840, "end_frame": 3859, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3789, "context_end_frame": 3908, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00192_ctx3789_3908_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00192_ctx3789_3908_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 192, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3840, "end_frame": 3859}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00192_ctx3789_3908_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00192_ctx3789_3908_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3840-3859\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3859, \"start_frame\": 3840}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:193", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3860, "end_frame": 3879, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3809, "context_end_frame": 3928, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00193_ctx3809_3928_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00193_ctx3809_3928_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 193, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3860, "end_frame": 3879}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00193_ctx3809_3928_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00193_ctx3809_3928_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3860-3879\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3879, \"start_frame\": 3860}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:194", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3880, "end_frame": 3899, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3829, "context_end_frame": 3948, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00194_ctx3829_3948_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00194_ctx3829_3948_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 194, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3880, "end_frame": 3899}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00194_ctx3829_3948_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00194_ctx3829_3948_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3880-3899\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3899, \"start_frame\": 3880}, \"next_action\": \"Pour liquid from white bottle\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:195", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3900, "end_frame": 3919, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3849, "context_end_frame": 3968, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00195_ctx3849_3968_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00195_ctx3849_3968_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 195, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3900, "end_frame": 3919}}, "label": "Pour liquid from white bottle", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00195_ctx3849_3968_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00195_ctx3849_3968_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3900-3919\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour liquid from white bottle\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3919, \"start_frame\": 3900}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:196", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3920, "end_frame": 3939, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3869, "context_end_frame": 3988, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00196_ctx3869_3988_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00196_ctx3869_3988_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 196, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3920, "end_frame": 3939}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00196_ctx3869_3988_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00196_ctx3869_3988_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3920-3939\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3939, \"start_frame\": 3920}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:197", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3940, "end_frame": 3959, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3889, "context_end_frame": 4008, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00197_ctx3889_4008_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00197_ctx3889_4008_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 197, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3940, "end_frame": 3959}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00197_ctx3889_4008_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00197_ctx3889_4008_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3940-3959\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3959, \"start_frame\": 3940}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:198", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3960, "end_frame": 3979, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3909, "context_end_frame": 4028, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00198_ctx3909_4028_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00198_ctx3909_4028_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 198, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3960, "end_frame": 3979}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00198_ctx3909_4028_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00198_ctx3909_4028_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3960-3979\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3979, \"start_frame\": 3960}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:199", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 3980, "end_frame": 3999, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3929, "context_end_frame": 4048, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00199_ctx3929_4048_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00199_ctx3929_4048_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 199, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3980, "end_frame": 3999}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00199_ctx3929_4048_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00199_ctx3929_4048_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 3980-3999\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 3999, \"start_frame\": 3980}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:200", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4000, "end_frame": 4019, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3949, "context_end_frame": 4068, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00200_ctx3949_4068_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00200_ctx3949_4068_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 200, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4000, "end_frame": 4019}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00200_ctx3949_4068_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00200_ctx3949_4068_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4000-4019\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4019, \"start_frame\": 4000}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\", \"mug\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:201", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4020, "end_frame": 4039, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3969, "context_end_frame": 4088, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00201_ctx3969_4088_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00201_ctx3969_4088_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 201, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4020, "end_frame": 4039}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00201_ctx3969_4088_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00201_ctx3969_4088_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4020-4039\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4039, \"start_frame\": 4020}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\", \"mug\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:202", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4040, "end_frame": 4059, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 3989, "context_end_frame": 4108, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00202_ctx3989_4108_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00202_ctx3989_4108_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 202, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4040, "end_frame": 4059}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00202_ctx3989_4108_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00202_ctx3989_4108_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4040-4059\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4059, \"start_frame\": 4040}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\", \"mug\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:203", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4060, "end_frame": 4079, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4009, "context_end_frame": 4128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00203_ctx4009_4128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00203_ctx4009_4128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 203, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4060, "end_frame": 4079}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00203_ctx4009_4128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00203_ctx4009_4128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4060-4079\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4079, \"start_frame\": 4060}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\", \"mug\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:204", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4080, "end_frame": 4099, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4029, "context_end_frame": 4148, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00204_ctx4029_4148_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00204_ctx4029_4148_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 204, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4080, "end_frame": 4099}}, "label": "Close bottle cap", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00204_ctx4029_4148_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00204_ctx4029_4148_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4080-4099\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Close bottle cap\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4099, \"start_frame\": 4080}, \"next_action\": \"Close bottle cap\", \"objects\": [\"white bottle\", \"coffee dripper\", \"scale\", \"mug\"], \"subtask\": \"Pour and close white bottle\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:220", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4400, "end_frame": 4419, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4349, "context_end_frame": 4468, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00220_ctx4349_4468_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00220_ctx4349_4468_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 220, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4400, "end_frame": 4419}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00220_ctx4349_4468_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00220_ctx4349_4468_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4400-4419\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4419, \"start_frame\": 4400}, \"next_action\": \"Place item on table\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:221", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4420, "end_frame": 4439, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4369, "context_end_frame": 4488, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00221_ctx4369_4488_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00221_ctx4369_4488_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 221, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4420, "end_frame": 4439}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00221_ctx4369_4488_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00221_ctx4369_4488_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4420-4439\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4439, \"start_frame\": 4420}, \"next_action\": \"Place item on table\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:222", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4440, "end_frame": 4459, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4389, "context_end_frame": 4508, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00222_ctx4389_4508_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00222_ctx4389_4508_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 222, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4440, "end_frame": 4459}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00222_ctx4389_4508_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00222_ctx4389_4508_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4440-4459\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4459, \"start_frame\": 4440}, \"next_action\": \"Place item on table\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:223", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4460, "end_frame": 4479, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4409, "context_end_frame": 4528, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00223_ctx4409_4528_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00223_ctx4409_4528_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 223, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4460, "end_frame": 4479}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00223_ctx4409_4528_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00223_ctx4409_4528_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4460-4479\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4479, \"start_frame\": 4460}, \"next_action\": \"Place item on table\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:224", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4480, "end_frame": 4499, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4429, "context_end_frame": 4548, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00224_ctx4429_4548_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00224_ctx4429_4548_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 224, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4480, "end_frame": 4499}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00224_ctx4429_4548_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00224_ctx4429_4548_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4480-4499\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4499, \"start_frame\": 4480}, \"next_action\": \"Place item on table\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:225", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4500, "end_frame": 4519, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4449, "context_end_frame": 4568, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00225_ctx4449_4568_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00225_ctx4449_4568_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 225, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4500, "end_frame": 4519}}, "label": "Place item on table", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00225_ctx4449_4568_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00225_ctx4449_4568_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4500-4519\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Place item on table\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4519, \"start_frame\": 4500}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:226", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4520, "end_frame": 4539, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4469, "context_end_frame": 4588, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00226_ctx4469_4588_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00226_ctx4469_4588_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 226, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4520, "end_frame": 4539}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00226_ctx4469_4588_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00226_ctx4469_4588_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4520-4539\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4539, \"start_frame\": 4520}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:227", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4540, "end_frame": 4559, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4489, "context_end_frame": 4608, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00227_ctx4489_4608_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00227_ctx4489_4608_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 227, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4540, "end_frame": 4559}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00227_ctx4489_4608_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00227_ctx4489_4608_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4540-4559\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4559, \"start_frame\": 4540}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:228", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4560, "end_frame": 4579, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4509, "context_end_frame": 4628, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00228_ctx4509_4628_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00228_ctx4509_4628_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 228, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4560, "end_frame": 4579}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00228_ctx4509_4628_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00228_ctx4509_4628_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4560-4579\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4579, \"start_frame\": 4560}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:229", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4580, "end_frame": 4599, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4529, "context_end_frame": 4648, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00229_ctx4529_4648_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00229_ctx4529_4648_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 229, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4580, "end_frame": 4599}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00229_ctx4529_4648_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00229_ctx4529_4648_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4580-4599\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4599, \"start_frame\": 4580}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:230", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4600, "end_frame": 4619, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4549, "context_end_frame": 4668, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00230_ctx4549_4668_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00230_ctx4549_4668_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 230, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4600, "end_frame": 4619}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00230_ctx4549_4668_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00230_ctx4549_4668_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4600-4619\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4619, \"start_frame\": 4600}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:231", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4620, "end_frame": 4639, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4569, "context_end_frame": 4688, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00231_ctx4569_4688_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00231_ctx4569_4688_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 231, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4620, "end_frame": 4639}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00231_ctx4569_4688_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00231_ctx4569_4688_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4620-4639\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4639, \"start_frame\": 4620}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:232", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4640, "end_frame": 4659, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4589, "context_end_frame": 4708, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00232_ctx4589_4708_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00232_ctx4589_4708_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 232, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4640, "end_frame": 4659}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00232_ctx4589_4708_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00232_ctx4589_4708_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4640-4659\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4659, \"start_frame\": 4640}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:233", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 4660, "end_frame": 4679, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4609, "context_end_frame": 4728, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00233_ctx4609_4728_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00233_ctx4609_4728_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 233, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4660, "end_frame": 4679}}, "label": "Wait/Prepare for pouring", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00233_ctx4609_4728_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00233_ctx4609_4728_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 4660-4679\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Wait/Prepare for pouring\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 4679, \"start_frame\": 4660}, \"next_action\": \"Wait/Prepare for pouring\", \"objects\": [\"coffee equipment\", \"small bottle\", \"white mug\", \"weighing scale\"], \"subtask\": \"Prepare for pouring\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:250", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5000, "end_frame": 5019, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4949, "context_end_frame": 5068, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00250_ctx4949_5068_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00250_ctx4949_5068_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 250, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5000, "end_frame": 5019}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00250_ctx4949_5068_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00250_ctx4949_5068_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5000-5019\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5019, \"start_frame\": 5000}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:251", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5020, "end_frame": 5039, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4969, "context_end_frame": 5088, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00251_ctx4969_5088_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00251_ctx4969_5088_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 251, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5020, "end_frame": 5039}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00251_ctx4969_5088_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00251_ctx4969_5088_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5020-5039\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5039, \"start_frame\": 5020}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:252", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5040, "end_frame": 5059, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 4989, "context_end_frame": 5108, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00252_ctx4989_5108_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00252_ctx4989_5108_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 252, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5040, "end_frame": 5059}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00252_ctx4989_5108_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00252_ctx4989_5108_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5040-5059\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5059, \"start_frame\": 5040}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:253", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5060, "end_frame": 5079, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5009, "context_end_frame": 5128, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00253_ctx5009_5128_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00253_ctx5009_5128_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 253, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5060, "end_frame": 5079}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00253_ctx5009_5128_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00253_ctx5009_5128_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5060-5079\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5079, \"start_frame\": 5060}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:254", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5080, "end_frame": 5099, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5029, "context_end_frame": 5148, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00254_ctx5029_5148_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00254_ctx5029_5148_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 254, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5080, "end_frame": 5099}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00254_ctx5029_5148_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00254_ctx5029_5148_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5080-5099\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5099, \"start_frame\": 5080}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:255", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5100, "end_frame": 5119, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5049, "context_end_frame": 5168, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00255_ctx5049_5168_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00255_ctx5049_5168_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 255, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5100, "end_frame": 5119}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00255_ctx5049_5168_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00255_ctx5049_5168_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5100-5119\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5119, \"start_frame\": 5100}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:256", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5120, "end_frame": 5139, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5069, "context_end_frame": 5188, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00256_ctx5069_5188_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00256_ctx5069_5188_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 256, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5120, "end_frame": 5139}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00256_ctx5069_5188_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00256_ctx5069_5188_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5120-5139\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5139, \"start_frame\": 5120}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:257", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5140, "end_frame": 5159, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5089, "context_end_frame": 5208, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00257_ctx5089_5208_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00257_ctx5089_5208_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 257, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5140, "end_frame": 5159}}, "label": "Pour coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00257_ctx5089_5208_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00257_ctx5089_5208_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5140-5159\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5159, \"start_frame\": 5140}, \"next_action\": \"Pour coffee\", \"objects\": [\"carafe\", \"coffee mug\", \"scale\"], \"subtask\": \"Pour coffee\", \"transition\": \"no\"}"}]} +{"id": "xperience-10m-sample:qa:290", "episode_id": "xperience-10m-sample", "split": "train", "target": "episode_qa", "prompt_type": "json_episode_understanding", "center_window": {"start_frame": 5800, "end_frame": 5819, "num_frames": 20}, "media": {"video_paths": [{"name": "fisheye_cam0.mp4", "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4"}], "context_start_frame": 5701, "context_end_frame": 5820, "max_video_frames": 16, "mosaic_video_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00290_ctx5701_5820_mosaic.mp4", "audio_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00290_ctx5701_5820_audio.wav"}, "sensor_feature_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/sensor_features/xperience-10m-sample_sensor_features.npz", "sensor_feature_index": 290, "sensor_feature_dim": 4262, "question": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.", "answer_json": {"action": "Pour milk into coffee", "subtask": "Pour milk into coffee", "objects": ["coffee mug", "stainless steel milk pitcher", "table", "milk bottle", "digital scale"], "contact": "yes", "transition": "no", "next_action": "Pour milk into coffee", "evidence_window": {"start_frame": 5800, "end_frame": 5819}}, "label": "Pour milk into coffee", "action_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "subtask_options": ["Handle gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Pick up and position kettle", "Pour and close white bottle", "Pour coffee", "Pour milk into coffee", "Prepare coffee equipment and scoop grounds", "Prepare for pouring", "Set down kettle and retrieve white bottle", "Transfer coffee grounds to dripper"], "label_options": ["Close bottle cap", "Grasp coffee scoop", "Grasp gooseneck kettle", "Hold coffee carafe", "Hold gooseneck kettle", "Lift gooseneck kettle", "Move kettle", "Move kettle away", "Pick up kettle", "Pick up white bottle", "Place item on table", "Place kettle on table", "Position kettle to pour", "Pour coffee", "Pour liquid from white bottle", "Pour milk into coffee", "Transfer coffee to dripper", "Wait/Prepare for pouring"], "messages": [{"role": "system", "content": [{"type": "text", "text": "You are an embodied episode-understanding model for Ropedia/Xperience-10M. Answer every question as strict JSON with these keys: action, subtask, objects, contact, transition, next_action, evidence_window. Use \"unknown\" when the evidence is missing instead of guessing."}]}, {"role": "user", "content": [{"type": "video", "video": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00290_ctx5701_5820_mosaic.mp4"}, {"type": "audio", "audio": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/media/xperience-10m-sample/xperience-10m-sample_w00290_ctx5701_5820_audio.wav"}, {"type": "text", "text": "Given the synchronized egocentric video/audio context and sensor window, identify the current embodied episode state.\nEpisode: xperience-10m-sample\nLabel window frames: 5800-5819\nReturn strict JSON only with keys: action, subtask, objects, contact, transition, next_action, evidence_window.\nUse \"unknown\" for fields that cannot be determined.\nKnown action labels:\n- Close bottle cap\n- Grasp coffee scoop\n- Grasp gooseneck kettle\n- Hold coffee carafe\n- Hold gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Move kettle away\n- Pick up kettle\n- Pick up white bottle\n- Place item on table\n- Place kettle on table\n- Position kettle to pour\n- Pour coffee\n- Pour liquid from white bottle\n- Pour milk into coffee\n- Transfer coffee to dripper\n- Wait/Prepare for pouring\nKnown subtask labels:\n- Handle gooseneck kettle\n- Lift gooseneck kettle\n- Move kettle\n- Pick up and position kettle\n- Pour and close white bottle\n- Pour coffee\n- Pour milk into coffee\n- Prepare coffee equipment and scoop grounds\n- Prepare for pouring\n- Set down kettle and retrieve white bottle\n- Transfer coffee grounds to dripper"}]}, {"role": "assistant", "content": "{\"action\": \"Pour milk into coffee\", \"contact\": \"yes\", \"evidence_window\": {\"end_frame\": 5819, \"start_frame\": 5800}, \"next_action\": \"Pour milk into coffee\", \"objects\": [\"coffee mug\", \"stainless steel milk pitcher\", \"table\", \"milk bottle\", \"digital scale\"], \"subtask\": \"Pour milk into coffee\", \"transition\": \"no\"}"}]} diff --git a/results/omni_finetune/dataset_manifest.json b/results/omni_finetune/dataset_manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..d6323794721189313c5103667e959f2ac24bbafe --- /dev/null +++ b/results/omni_finetune/dataset_manifest.json @@ -0,0 +1,175 @@ +{ + "run_id": "xperience10m_qwen3_omni_32ep_dataset", + "dataset_path": "/path/to/ropedia_workspace/ropedia-episode-task-suite/results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "num_samples": 128, + "num_episodes": 1, + "split_counts": { + "train": 128 + }, + "label_counts": { + "Close bottle cap": 9, + "Pick up kettle": 8, + "Position kettle to pour": 8, + "Move kettle": 8, + "Hold coffee carafe": 8, + "Grasp coffee scoop": 8, + "Transfer coffee to dripper": 8, + "Hold gooseneck kettle": 8, + "Lift gooseneck kettle": 8, + "Move kettle away": 8, + "Wait/Prepare for pouring": 8, + "Pour coffee": 8, + "Grasp gooseneck kettle": 6, + "Place kettle on table": 6, + "Pick up white bottle": 6, + "Pour liquid from white bottle": 6, + "Place item on table": 6, + "Pour milk into coffee": 1 + }, + "action_options": [ + "Close bottle cap", + "Grasp coffee scoop", + "Grasp gooseneck kettle", + "Hold coffee carafe", + "Hold gooseneck kettle", + "Lift gooseneck kettle", + "Move kettle", + "Move kettle away", + "Pick up kettle", + "Pick up white bottle", + "Place item on table", + "Place kettle on table", + "Position kettle to pour", + "Pour coffee", + "Pour liquid from white bottle", + "Pour milk into coffee", + "Transfer coffee to dripper", + "Wait/Prepare for pouring" + ], + "subtask_options": [ + "Handle gooseneck kettle", + "Lift gooseneck kettle", + "Move kettle", + "Pick up and position kettle", + "Pour and close white bottle", + "Pour coffee", + "Pour milk into coffee", + "Prepare coffee equipment and scoop grounds", + "Prepare for pouring", + "Set down kettle and retrieve white bottle", + "Transfer coffee grounds to dripper" + ], + "clip_policy": { + "label_window_frames": 20, + "qwen_context_frames": 120, + "max_video_frames": 16, + "audio_span": "same_as_video_context", + "mosaic": "2x3 multi-camera grid" + }, + "feature_manifest": [ + { + "name": "hand_left_joints", + "start": 0, + "end": 441, + "dim": 441 + }, + { + "name": "hand_right_joints", + "start": 441, + "end": 882, + "dim": 441 + }, + { + "name": "body_joints", + "start": 882, + "end": 1974, + "dim": 1092 + }, + { + "name": "body_contacts", + "start": 1974, + "end": 2121, + "dim": 147 + }, + { + "name": "camera_translation", + "start": 2121, + "end": 2142, + "dim": 21 + }, + { + "name": "camera_rotation_matrix", + "start": 2142, + "end": 2205, + "dim": 63 + }, + { + "name": "imu_accel_gyro", + "start": 2205, + "end": 2247, + "dim": 42 + }, + { + "name": "depth_confidence", + "start": 2247, + "end": 3227, + "dim": 980 + }, + { + "name": "caption_objects_interaction_text", + "start": 3227, + "end": 4123, + "dim": 896 + }, + { + "name": "slam_point_cloud", + "start": 4123, + "end": 4145, + "dim": 22 + }, + { + "name": "calibration", + "start": 4145, + "end": 4262, + "dim": 117 + } + ], + "available_modalities": [ + { + "episode_id": "xperience-10m-sample", + "modalities": [ + { + "modality": "depth_confidence", + "shape": [ + 5821, + 140 + ] + }, + { + "modality": "caption_text", + "shape": [ + 5821, + 128 + ], + "fields": "objects,interaction" + }, + { + "modality": "slam_point_cloud_static", + "shape": [ + 22 + ] + }, + { + "modality": "calibration_static", + "shape": [ + 117 + ] + } + ] + } + ], + "notes": [ + "Assistant answers are strict JSON for episode understanding, not robot-control policies.", + "Sensor features are stored as NPZ pointers; raw annotation.hdf5 is not copied into the dataset records." + ] +} \ No newline at end of file diff --git a/results/omni_finetune/episode_manifest.json b/results/omni_finetune/episode_manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..8de50aa7fb60a0372b1aff262a2f8a5b6e6d5289 --- /dev/null +++ b/results/omni_finetune/episode_manifest.json @@ -0,0 +1,219 @@ +{ + "summary": { + "num_episodes": 1, + "total_bytes": 2021338279, + "train_minimal_bytes": 2021338279, + "split_counts": { + "train": 1 + }, + "split_fractions": { + "train": 0.8, + "val": 0.0, + "test": 0.2, + "seed": 7 + }, + "windowing": { + "window_frames": 20, + "stride_frames": 20, + "min_label_fraction": 0.6 + }, + "notes": [ + "train_minimal_bytes excludes visualization.rrd because model training does not need it.", + "This file is metadata-only; it does not copy or download raw data.", + "Splits are assigned by whole episode to avoid window leakage." + ] + }, + "episodes": [ + { + "episode_id": "xperience-10m-sample", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample", + "annotation": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/annotation.hdf5", + "frame_count": 5821, + "main_task": "Making pour-over coffee", + "files": [ + { + "name": "annotation.hdf5", + "bytes": 1931496028, + "exists": true + }, + { + "name": "fisheye_cam0.mp4", + "bytes": 89842251, + "exists": true + }, + { + "name": "fisheye_cam1.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "fisheye_cam2.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "fisheye_cam3.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "stereo_left.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "stereo_right.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "visualization.rrd", + "bytes": 0, + "exists": false + } + ], + "videos": [ + { + "name": "fisheye_cam0.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam0.mp4", + "bytes": 89842251, + "exists": true + }, + { + "name": "fisheye_cam1.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam1.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "fisheye_cam2.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam2.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "fisheye_cam3.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/fisheye_cam3.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "stereo_left.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/stereo_left.mp4", + "bytes": 0, + "exists": false + }, + { + "name": "stereo_right.mp4", + "path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample/stereo_right.mp4", + "bytes": 0, + "exists": false + } + ], + "hdf5_modalities": { + "calibration": true, + "slam_pose": true, + "slam_point_cloud": true, + "depth": true, + "depth_confidence": true, + "hand_mocap": true, + "body_mocap": true, + "contacts": true, + "imu": true, + "caption": true, + "captions": false + }, + "label_stats": { + "main_task": "Making pour-over coffee", + "segments": 18, + "frame_labels": { + "action": { + "Hold gooseneck kettle": 800, + "Pour coffee": 800, + "Position kettle to pour": 640, + "Lift gooseneck kettle": 480, + "Close bottle cap": 480, + "Wait/Prepare for pouring": 480, + "Transfer coffee to dripper": 439, + "Grasp coffee scoop": 321, + "Hold coffee carafe": 240, + "Move kettle": 200, + "Pick up kettle": 160, + "Move kettle away": 160, + "Grasp gooseneck kettle": 120, + "Place kettle on table": 120, + "Pick up white bottle": 120, + "Pour liquid from white bottle": 120, + "Place item on table": 120, + "Pour milk into coffee": 21 + }, + "subtask": { + "Handle gooseneck kettle": 839, + "Pick up and position kettle": 761, + "Pour coffee": 761, + "Prepare for pouring": 600, + "Prepare coffee equipment and scoop grounds": 561, + "Lift gooseneck kettle": 561, + "Pour and close white bottle": 561, + "Transfer coffee grounds to dripper": 400, + "Set down kettle and retrieve white bottle": 400, + "Move kettle": 200, + "Pour milk into coffee": 60, + "Position kettle to pour": 39, + "Secure coffee container": 39, + "Move bottle to coffee equipment": 39 + } + }, + "window_labels": { + "action": { + "Hold gooseneck kettle": 40, + "Pour coffee": 40, + "Position kettle to pour": 32, + "Lift gooseneck kettle": 24, + "Close bottle cap": 24, + "Wait/Prepare for pouring": 24, + "Transfer coffee to dripper": 22, + "Grasp coffee scoop": 16, + "Hold coffee carafe": 12, + "Move kettle": 10, + "Pick up kettle": 8, + "Move kettle away": 8, + "Grasp gooseneck kettle": 6, + "Place kettle on table": 6, + "Pick up white bottle": 6, + "Pour liquid from white bottle": 6, + "Place item on table": 6, + "Pour milk into coffee": 1 + }, + "subtask": { + "Handle gooseneck kettle": 42, + "Pick up and position kettle": 38, + "Pour coffee": 38, + "Prepare for pouring": 30, + "Prepare coffee equipment and scoop grounds": 28, + "Lift gooseneck kettle": 28, + "Pour and close white bottle": 28, + "Transfer coffee grounds to dripper": 20, + "Set down kettle and retrieve white bottle": 20, + "Move kettle": 10, + "Pour milk into coffee": 3, + "Position kettle to pour": 2, + "Secure coffee container": 2, + "Move bottle to coffee equipment": 2 + } + }, + "num_labeled_windows": { + "action": 291, + "subtask": 291 + } + }, + "total_bytes": 2021338279, + "train_minimal_bytes": 2021338279, + "has_annotation": true, + "has_any_video": true, + "has_all_videos": false, + "has_rrd": false, + "split": "train" + } + ] +} \ No newline at end of file diff --git a/results/omni_finetune/hf_upload/README.md b/results/omni_finetune/hf_upload/README.md new file mode 100644 index 0000000000000000000000000000000000000000..47aa853c5b16380d4760a2fcac6fbc9ad5c76fc0 --- /dev/null +++ b/results/omni_finetune/hf_upload/README.md @@ -0,0 +1,61 @@ +--- +license: other +base_model: Qwen/Qwen3-Omni-30B-A3B-Instruct +library_name: peft +tags: + - robotics + - embodied-ai + - multimodal + - xperience-10m + - qwen3-omni + - lora + - readiness-check +datasets: + - ropedia-ai/xperience-10m +pipeline_tag: image-text-to-text +--- + +# Xperience-10M Qwen3-Omni LoRA (pilot artifact) + +## What this is +This repository contains a Qwen3-Omni LoRA adapter produced from an initial end-to-end +technical readiness run on Xperience-10M data. + +It is a **readiness artifact** from one validated training run with: + +- Backbone: `Qwen/Qwen3-Omni-30B-A3B-Instruct` +- Adapter: LoRA, rank=16, alpha=32, dropout=0.05 +- Data source: `/path/to/ropedia_workspace/modelscope_data` (single available episode set at run time) +- Windows used: `128` train windows +- Processes: `8` +- Train split episode-level leakage control: **not a full 32-episode run yet** + +## Current Scale-Up Status + +The real 32-episode pilot is prepared but not complete: + +- staging workflow: configured +- Selection: 32 complete episodes from 32 different session UUIDs +- Estimated raw subset: about 72 GB, excluding `visualization.rrd` +- Blocker: full `ropedia-ai/xperience-10m` access is still pending Hugging Face gated approval +- Claim boundary: no 32-episode held-out metrics are claimed until multi-episode data access, manifest building, training, and evaluation finish + +## Files +- `README.md` — this file +- `adapter_config.json` — LoRA configuration metadata +- `adapter_model.safetensors` — LoRA checkpoint weights +- `training_metadata.json` — run metadata and hyperparameters +- `processor_config.json`, `tokenizer.json`, `tokenizer_config.json`, `chat_template.jinja` + +## Reproducibility note +This artifact is intentionally labeled as a pilot artifact. It should be treated as: + +1. A proof of successful Qwen3-Omni pipeline wiring +2. Not a final benchmark model for Xperience-10M full 32-episode fine-tuning +3. A starting point for the next run once full pilot data are available + +## Source +Project and task definitions are documented in: +- `https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite` +- `https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/` +- https://cy0307-ropedia-xperience-10m-task-suite.static.hf.space/ diff --git a/results/omni_finetune/hf_upload/adapter_config.json b/results/omni_finetune/hf_upload/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..ef8ca4226f1b89d8b01f981e04775fb87c7e0543 --- /dev/null +++ b/results/omni_finetune/hf_upload/adapter_config.json @@ -0,0 +1,49 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": { + "base_model_class": "Qwen3OmniMoeThinkerForConditionalGeneration", + "parent_library": "transformers.models.qwen3_omni_moe.modeling_qwen3_omni_moe" + }, + "base_model_name_or_path": "", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "peft_version": "0.18.1", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "up_proj", + "q_proj", + "v_proj", + "k_proj", + "down_proj", + "o_proj" + ], + "target_parameters": null, + "task_type": null, + "trainable_token_indices": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/results/omni_finetune/hf_upload/chat_template.jinja b/results/omni_finetune/hf_upload/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..1efcce1402ae5d1d61d4a4cf56accea5921251cf --- /dev/null +++ b/results/omni_finetune/hf_upload/chat_template.jinja @@ -0,0 +1,122 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {{- "<|vision_start|><|image_pad|><|vision_end|>" }} + {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %} + {{- "<|audio_start|><|audio_pad|><|audio_end|>" }} + {%- elif content.type == 'video' or 'video' in content %} + {{- "<|vision_start|><|video_pad|><|vision_end|>" }} + {%- elif content.type == 'text' %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- endif %} + {{- '\n\n' }} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }} + {%- else %} + {%- for content in messages[0].content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {{- '<|im_start|>system\n' +"<|vision_start|><|image_pad|><|vision_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'audio' or 'audio' in content or 'audio_url' in content %} + {{- '<|im_start|>system\n' +"<|audio_start|><|audio_pad|><|audio_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'video' or 'video' in content %} + {{- '<|im_start|>system\n' +"<|vision_start|><|video_pad|><|vision_end|>"+ '<|im_end|>\n' }} + {%- elif content.type == 'text' %} + {{- '<|im_start|>system\n' +content.text+ '<|im_end|>\n' }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} +{%- endfor %} +{%- for message in messages %} + {%- if message.content is string %} + {%- set content = message.content %} + {%- else %} + {%- set content = namespace(text="") %} + {%- for mcontent in message.content %} + {%- if mcontent.type == 'image' or 'image' in mcontent or 'image_url' in mcontent %} + {%- set content.text = content.text~"<|vision_start|><|image_pad|><|vision_end|>" %} + {%- elif mcontent.type == 'audio' or 'audio' in mcontent or 'audio_url' in mcontent %} + {%- set content.text = content.text~"<|audio_start|><|audio_pad|><|audio_end|>" %} + {%- elif mcontent.type == 'video' or 'video' in mcontent %} + {%- set content.text = content.text~"<|vision_start|><|video_pad|><|vision_end|>" %} + {%- elif mcontent.type == 'text' %} + {%- set content.text = content.text~mcontent.text %} + {%- endif %} + {%- endfor %} + {%- set content = content.text %} + {%- endif %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = "" %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- if loop.index0 > ns.last_query_index %} + {%- if loop.last or (not loop.last and reasoning_content) %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip("\n") + '\n\n\n' + content.lstrip('\n') }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and content) or (not loop.first) %}{{- '\n' }}{%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}{{- '<|im_start|>user' }}{%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}{{- '<|im_end|>\n' }}{%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %}{{- '\n\n\n\n' }}{%- endif %} +{%- endif %} \ No newline at end of file diff --git a/results/omni_finetune/hf_upload/processor_config.json b/results/omni_finetune/hf_upload/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..f5e20e1e0b4ab8f002e2475fc2d49e8910d5c627 --- /dev/null +++ b/results/omni_finetune/hf_upload/processor_config.json @@ -0,0 +1,114 @@ +{ + "feature_extractor": { + "chunk_length": 30, + "dither": 0.0, + "feature_extractor_type": "WhisperFeatureExtractor", + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_pixels": 12845056, + "merge_size": 2, + "min_pixels": 3136, + "n_fft": 400, + "n_samples": 480000, + "nb_max_frames": 3000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "return_attention_mask": true, + "sampling_rate": 16000, + "temporal_patch_size": 2 + }, + "image_processor": { + "data_format": "channels_first", + "dither": 0.0, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "n_fft": 400, + "n_samples": 4800000, + "nb_max_frames": 30000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_attention_mask": true, + "sampling_rate": 16000, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3OmniMoeProcessor", + "video_processor": { + "data_format": "channels_first", + "default_to_square": true, + "dither": 0.0, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": false, + "feature_extractor_type": "WhisperFeatureExtractor", + "feature_size": 128, + "hop_length": 160, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "n_fft": 400, + "n_samples": 4800000, + "nb_max_frames": 30000, + "padding_side": "right", + "padding_value": 0.0, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_attention_mask": true, + "return_metadata": false, + "sampling_rate": 16000, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen2VLVideoProcessor" + } +} diff --git a/results/omni_finetune/hf_upload/tokenizer.json b/results/omni_finetune/hf_upload/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f7ab87da8ac9fbb6ebdd03fb50597a92933ce3c6 --- /dev/null +++ b/results/omni_finetune/hf_upload/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d41145c408f3062e824f965be1c29854cd809e111cdf8170aa8b0bcd5d1fab +size 11424262 diff --git a/results/omni_finetune/hf_upload/tokenizer_config.json b/results/omni_finetune/hf_upload/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..22039d0853e5588d2b04702b4dbde3ff8546edfc --- /dev/null +++ b/results/omni_finetune/hf_upload/tokenizer_config.json @@ -0,0 +1,52 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>", + "<|audio_start|>", + "<|audio_end|>", + "", + "", + "", + "<|audio_pad|>" + ], + "image_token": "<|image_pad|>", + "is_local": true, + "model_max_length": 131072, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3OmniMoeProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/results/omni_finetune/hf_upload/training_metadata.json b/results/omni_finetune/hf_upload/training_metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..50814ee383d9cf91fe31fbdd5e76a2cdacff9215 --- /dev/null +++ b/results/omni_finetune/hf_upload/training_metadata.json @@ -0,0 +1,32 @@ +{ + "run_id": "xperience10m_qwen3_omni_32ep_lora", + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "num_processes": 8, + "num_train_samples": 128, + "num_val_samples": 0, + "history": [ + { + "epoch": 1, + "train_loss": 10.936363816261292, + "val_loss": null, + "global_step": 16 + } + ], + "lora": { + "r": 16, + "alpha": 32, + "dropout": 0.05, + "target_modules": [ + "q_proj", + "k_proj", + "v_proj", + "o_proj", + "gate_proj", + "up_proj", + "down_proj" + ] + }, + "use_audio_in_video": false +} \ No newline at end of file diff --git a/results/omni_finetune/lora_config.yaml b/results/omni_finetune/lora_config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..22f96748a9d9ec2cc8bd8f73d5bad1f6851ef5d3 --- /dev/null +++ b/results/omni_finetune/lora_config.yaml @@ -0,0 +1,10 @@ +run_id: xperience10m_qwen3_omni_32ep_lora +stage: qwen_lora_text_video_audio +model_id: /path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct +dataset_jsonl: results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl +checkpoint_dir: /path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora +num_processes: 8 +epochs: 1 +learning_rate: 0.0001 +lora_r: 16 +lora_alpha: 32 diff --git a/results/omni_finetune/metrics.json b/results/omni_finetune/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..d08c4b27d7045396b9cab00581430b830c044f1f --- /dev/null +++ b/results/omni_finetune/metrics.json @@ -0,0 +1,68 @@ +{ + "num_samples": 128, + "accuracy": 0.0, + "macro_f1": 0.0, + "labels": [ + "Close bottle cap", + "Grasp coffee scoop", + "Grasp gooseneck kettle", + "Hold coffee carafe", + "Hold gooseneck kettle", + "Lift gooseneck kettle", + "Move kettle", + "Move kettle away", + "Pick up kettle", + "Pick up white bottle", + "Place item on table", + "Place kettle on table", + "Position kettle to pour", + "Pour coffee", + "Pour liquid from white bottle", + "Pour milk into coffee", + "Transfer coffee to dripper", + "Wait/Prepare for pouring", + "unknown" + ], + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "adapter_dir": "checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "eval_split": "train", + "train_split": "train", + "num_eval_episodes": 1, + "unseen_eval_labels": [], + "num_unseen_label_samples": 0, + "seen_label_accuracy": 0.0, + "unseen_label_accuracy": null, + "eval_label_counts": { + "Close bottle cap": 9, + "Pick up kettle": 8, + "Position kettle to pour": 8, + "Move kettle": 8, + "Hold coffee carafe": 8, + "Grasp coffee scoop": 8, + "Transfer coffee to dripper": 8, + "Hold gooseneck kettle": 8, + "Lift gooseneck kettle": 8, + "Move kettle away": 8, + "Wait/Prepare for pouring": 8, + "Pour coffee": 8, + "Grasp gooseneck kettle": 6, + "Place kettle on table": 6, + "Pick up white bottle": 6, + "Pour liquid from white bottle": 6, + "Place item on table": 6, + "Pour milk into coffee": 1 + }, + "json_validity_rate": 1.0, + "action_macro_f1": 0.0, + "subtask_accuracy": 0.015625, + "transition_accuracy": 0.0, + "next_action_accuracy": 0.0078125, + "contact_accuracy": 0.0, + "object_micro_f1": 0.031496062992125984, + "caption_window_grounding": { + "mrr": null, + "recall_at_5": null, + "note": "Grounding ranking requires a retrieval candidate set; JSON evidence_window is stored for later scoring." + } +} \ No newline at end of file diff --git a/results/omni_finetune/metrics_eval.json b/results/omni_finetune/metrics_eval.json new file mode 100644 index 0000000000000000000000000000000000000000..d08c4b27d7045396b9cab00581430b830c044f1f --- /dev/null +++ b/results/omni_finetune/metrics_eval.json @@ -0,0 +1,68 @@ +{ + "num_samples": 128, + "accuracy": 0.0, + "macro_f1": 0.0, + "labels": [ + "Close bottle cap", + "Grasp coffee scoop", + "Grasp gooseneck kettle", + "Hold coffee carafe", + "Hold gooseneck kettle", + "Lift gooseneck kettle", + "Move kettle", + "Move kettle away", + "Pick up kettle", + "Pick up white bottle", + "Place item on table", + "Place kettle on table", + "Position kettle to pour", + "Pour coffee", + "Pour liquid from white bottle", + "Pour milk into coffee", + "Transfer coffee to dripper", + "Wait/Prepare for pouring", + "unknown" + ], + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "adapter_dir": "checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "eval_split": "train", + "train_split": "train", + "num_eval_episodes": 1, + "unseen_eval_labels": [], + "num_unseen_label_samples": 0, + "seen_label_accuracy": 0.0, + "unseen_label_accuracy": null, + "eval_label_counts": { + "Close bottle cap": 9, + "Pick up kettle": 8, + "Position kettle to pour": 8, + "Move kettle": 8, + "Hold coffee carafe": 8, + "Grasp coffee scoop": 8, + "Transfer coffee to dripper": 8, + "Hold gooseneck kettle": 8, + "Lift gooseneck kettle": 8, + "Move kettle away": 8, + "Wait/Prepare for pouring": 8, + "Pour coffee": 8, + "Grasp gooseneck kettle": 6, + "Place kettle on table": 6, + "Pick up white bottle": 6, + "Pour liquid from white bottle": 6, + "Place item on table": 6, + "Pour milk into coffee": 1 + }, + "json_validity_rate": 1.0, + "action_macro_f1": 0.0, + "subtask_accuracy": 0.015625, + "transition_accuracy": 0.0, + "next_action_accuracy": 0.0078125, + "contact_accuracy": 0.0, + "object_micro_f1": 0.031496062992125984, + "caption_window_grounding": { + "mrr": null, + "recall_at_5": null, + "note": "Grounding ranking requires a retrieval candidate set; JSON evidence_window is stored for later scoring." + } +} \ No newline at end of file diff --git a/results/omni_finetune/per_class_metrics_eval.csv b/results/omni_finetune/per_class_metrics_eval.csv new file mode 100644 index 0000000000000000000000000000000000000000..0d77ab9bf9c9ab4e7ee4de2c8e73f656f345a7cc --- /dev/null +++ b/results/omni_finetune/per_class_metrics_eval.csv @@ -0,0 +1,20 @@ +class_name,support,predicted,precision,recall,f1 +Close bottle cap,9,0,0.0,0.0,0.0 +Grasp coffee scoop,8,0,0.0,0.0,0.0 +Grasp gooseneck kettle,6,0,0.0,0.0,0.0 +Hold coffee carafe,8,0,0.0,0.0,0.0 +Hold gooseneck kettle,8,0,0.0,0.0,0.0 +Lift gooseneck kettle,8,0,0.0,0.0,0.0 +Move kettle,8,0,0.0,0.0,0.0 +Move kettle away,8,0,0.0,0.0,0.0 +Pick up kettle,8,0,0.0,0.0,0.0 +Pick up white bottle,6,0,0.0,0.0,0.0 +Place item on table,6,0,0.0,0.0,0.0 +Place kettle on table,6,0,0.0,0.0,0.0 +Position kettle to pour,8,0,0.0,0.0,0.0 +Pour coffee,8,0,0.0,0.0,0.0 +Pour liquid from white bottle,6,0,0.0,0.0,0.0 +Pour milk into coffee,1,1,0.0,0.0,0.0 +Transfer coffee to dripper,8,0,0.0,0.0,0.0 +Wait/Prepare for pouring,8,40,0.0,0.0,0.0 +unknown,0,87,0.0,0.0,0.0 diff --git a/results/omni_finetune/pipeline_status.jsonl b/results/omni_finetune/pipeline_status.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..a39dd026e5ddf008550887fdc6ea11c5bc470977 --- /dev/null +++ b/results/omni_finetune/pipeline_status.jsonl @@ -0,0 +1,6 @@ +{"event": "phase", "phase": "preflight", "timestamp": 1780144750.6292024} +{"event": "phase", "phase": "download_qwen3_omni_instruct", "timestamp": 1780144757.9026744} +{"event": "phase", "phase": "build_manifest", "timestamp": 1780144815.232617} +{"event": "phase", "phase": "export_dataset", "timestamp": 1780144815.5557294} +{"event": "phase", "phase": "qwen_zero_shot_smoke", "timestamp": 1780144955.9370437} +{"event": "phase", "phase": "train_8gpu_lora", "timestamp": 1780144963.14272} diff --git a/results/omni_finetune/predictions_eval.csv b/results/omni_finetune/predictions_eval.csv new file mode 100644 index 0000000000000000000000000000000000000000..b14335228f48ccdb63ff5668dcb261ea7985230d --- /dev/null +++ b/results/omni_finetune/predictions_eval.csv @@ -0,0 +1,1510 @@ +id,target,split,episode_id,center_window,true_label,raw_prediction,predicted_label,correct,true_label_seen_in_train +xperience-10m-sample:qa:0,episode_qa,train,xperience-10m-sample,"{'start_frame': 0, 'end_frame': 19, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""0-19"" +} +```",unknown,0,1 +xperience-10m-sample:qa:1,episode_qa,train,xperience-10m-sample,"{'start_frame': 20, 'end_frame': 39, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:2,episode_qa,train,xperience-10m-sample,"{'start_frame': 40, 'end_frame': 59, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""40-59"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:3,episode_qa,train,xperience-10m-sample,"{'start_frame': 60, 'end_frame': 79, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""60-79"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:4,episode_qa,train,xperience-10m-sample,"{'start_frame': 80, 'end_frame': 99, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""80-99"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:5,episode_qa,train,xperience-10m-sample,"{'start_frame': 100, 'end_frame': 119, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""100-119"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:6,episode_qa,train,xperience-10m-sample,"{'start_frame': 120, 'end_frame': 139, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""120-139"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:7,episode_qa,train,xperience-10m-sample,"{'start_frame': 140, 'end_frame': 159, 'num_frames': 20}",Pick up kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""140-159"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:8,episode_qa,train,xperience-10m-sample,"{'start_frame': 160, 'end_frame': 179, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""160-179"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:9,episode_qa,train,xperience-10m-sample,"{'start_frame': 180, 'end_frame': 199, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""180-199"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:10,episode_qa,train,xperience-10m-sample,"{'start_frame': 200, 'end_frame': 219, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""200-219"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:11,episode_qa,train,xperience-10m-sample,"{'start_frame': 220, 'end_frame': 239, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""220-239"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:12,episode_qa,train,xperience-10m-sample,"{'start_frame': 240, 'end_frame': 259, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""240-259"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:13,episode_qa,train,xperience-10m-sample,"{'start_frame': 260, 'end_frame': 279, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""260-279"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:14,episode_qa,train,xperience-10m-sample,"{'start_frame': 280, 'end_frame': 299, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""280-299"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:15,episode_qa,train,xperience-10m-sample,"{'start_frame': 300, 'end_frame': 319, 'num_frames': 20}",Position kettle to pour,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:40,episode_qa,train,xperience-10m-sample,"{'start_frame': 800, 'end_frame': 819, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""800-819"" +} +```",unknown,0,1 +xperience-10m-sample:qa:41,episode_qa,train,xperience-10m-sample,"{'start_frame': 820, 'end_frame': 839, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""820-839"" +} +```",unknown,0,1 +xperience-10m-sample:qa:42,episode_qa,train,xperience-10m-sample,"{'start_frame': 840, 'end_frame': 859, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:43,episode_qa,train,xperience-10m-sample,"{'start_frame': 860, 'end_frame': 879, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:44,episode_qa,train,xperience-10m-sample,"{'start_frame': 880, 'end_frame': 899, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""880-899"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:45,episode_qa,train,xperience-10m-sample,"{'start_frame': 900, 'end_frame': 919, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""900-919"" +} +```",unknown,0,1 +xperience-10m-sample:qa:46,episode_qa,train,xperience-10m-sample,"{'start_frame': 920, 'end_frame': 939, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""920-939"" +} +```",unknown,0,1 +xperience-10m-sample:qa:47,episode_qa,train,xperience-10m-sample,"{'start_frame': 940, 'end_frame': 959, 'num_frames': 20}",Move kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""940-959"" +} +```",unknown,0,1 +xperience-10m-sample:qa:50,episode_qa,train,xperience-10m-sample,"{'start_frame': 1000, 'end_frame': 1019, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1000-1019"" +} +```",unknown,0,1 +xperience-10m-sample:qa:51,episode_qa,train,xperience-10m-sample,"{'start_frame': 1020, 'end_frame': 1039, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1020-1039"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:52,episode_qa,train,xperience-10m-sample,"{'start_frame': 1040, 'end_frame': 1059, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:53,episode_qa,train,xperience-10m-sample,"{'start_frame': 1060, 'end_frame': 1079, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1060-1079"" +} +```",unknown,0,1 +xperience-10m-sample:qa:54,episode_qa,train,xperience-10m-sample,"{'start_frame': 1080, 'end_frame': 1099, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1080-1099"" +} +```",unknown,0,1 +xperience-10m-sample:qa:55,episode_qa,train,xperience-10m-sample,"{'start_frame': 1100, 'end_frame': 1119, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1100-1119"" +} +```",unknown,0,1 +xperience-10m-sample:qa:56,episode_qa,train,xperience-10m-sample,"{'start_frame': 1120, 'end_frame': 1139, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1120-1139"" +} +```",unknown,0,1 +xperience-10m-sample:qa:57,episode_qa,train,xperience-10m-sample,"{'start_frame': 1140, 'end_frame': 1159, 'num_frames': 20}",Hold coffee carafe,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1140-1159"" +} +```",unknown,0,1 +xperience-10m-sample:qa:62,episode_qa,train,xperience-10m-sample,"{'start_frame': 1240, 'end_frame': 1259, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [""gooseneck kettle""], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1240-1259"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:63,episode_qa,train,xperience-10m-sample,"{'start_frame': 1260, 'end_frame': 1279, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1260-1279"" +} +```",unknown,0,1 +xperience-10m-sample:qa:64,episode_qa,train,xperience-10m-sample,"{'start_frame': 1280, 'end_frame': 1299, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1280-1299"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:65,episode_qa,train,xperience-10m-sample,"{'start_frame': 1300, 'end_frame': 1319, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1300-1319"" +} +```",unknown,0,1 +xperience-10m-sample:qa:66,episode_qa,train,xperience-10m-sample,"{'start_frame': 1320, 'end_frame': 1339, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:67,episode_qa,train,xperience-10m-sample,"{'start_frame': 1340, 'end_frame': 1359, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1340-1359"" +} +```",unknown,0,1 +xperience-10m-sample:qa:68,episode_qa,train,xperience-10m-sample,"{'start_frame': 1360, 'end_frame': 1379, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1360-1379"" +} +```",unknown,0,1 +xperience-10m-sample:qa:69,episode_qa,train,xperience-10m-sample,"{'start_frame': 1380, 'end_frame': 1399, 'num_frames': 20}",Grasp coffee scoop,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1380-1399"" +} +```",unknown,0,1 +xperience-10m-sample:qa:78,episode_qa,train,xperience-10m-sample,"{'start_frame': 1560, 'end_frame': 1579, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1560-1579"" +} +```",unknown,0,1 +xperience-10m-sample:qa:79,episode_qa,train,xperience-10m-sample,"{'start_frame': 1580, 'end_frame': 1599, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1580-1599"" +} +```",unknown,0,1 +xperience-10m-sample:qa:80,episode_qa,train,xperience-10m-sample,"{'start_frame': 1600, 'end_frame': 1619, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1600-1619"" +} +```",unknown,0,1 +xperience-10m-sample:qa:81,episode_qa,train,xperience-10m-sample,"{'start_frame': 1620, 'end_frame': 1639, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1620-1639"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:82,episode_qa,train,xperience-10m-sample,"{'start_frame': 1640, 'end_frame': 1659, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1640-1659"" +} +```",unknown,0,1 +xperience-10m-sample:qa:83,episode_qa,train,xperience-10m-sample,"{'start_frame': 1660, 'end_frame': 1679, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1660-1679"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:84,episode_qa,train,xperience-10m-sample,"{'start_frame': 1680, 'end_frame': 1699, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""1680-1699"" +} +```",unknown,0,1 +xperience-10m-sample:qa:85,episode_qa,train,xperience-10m-sample,"{'start_frame': 1700, 'end_frame': 1719, 'num_frames': 20}",Transfer coffee to dripper,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""1700-1719"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:100,episode_qa,train,xperience-10m-sample,"{'start_frame': 2000, 'end_frame': 2019, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2000-2019"" +} +```",unknown,0,1 +xperience-10m-sample:qa:101,episode_qa,train,xperience-10m-sample,"{'start_frame': 2020, 'end_frame': 2039, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2020-2039"" +} +```",unknown,0,1 +xperience-10m-sample:qa:102,episode_qa,train,xperience-10m-sample,"{'start_frame': 2040, 'end_frame': 2059, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2040-2059"" +} +```",unknown,0,1 +xperience-10m-sample:qa:103,episode_qa,train,xperience-10m-sample,"{'start_frame': 2060, 'end_frame': 2079, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2060-2079"" +} +```",unknown,0,1 +xperience-10m-sample:qa:104,episode_qa,train,xperience-10m-sample,"{'start_frame': 2080, 'end_frame': 2099, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2080-2099"" +} +```",unknown,0,1 +xperience-10m-sample:qa:105,episode_qa,train,xperience-10m-sample,"{'start_frame': 2100, 'end_frame': 2119, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2100-2119"" +} +```",unknown,0,1 +xperience-10m-sample:qa:106,episode_qa,train,xperience-10m-sample,"{'start_frame': 2120, 'end_frame': 2139, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""unknown"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2120-2139"" +} +```",unknown,0,1 +xperience-10m-sample:qa:107,episode_qa,train,xperience-10m-sample,"{'start_frame': 2140, 'end_frame': 2159, 'num_frames': 20}",Hold gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2140-2159"" +} +```",unknown,0,1 +xperience-10m-sample:qa:140,episode_qa,train,xperience-10m-sample,"{'start_frame': 2800, 'end_frame': 2819, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2800-2819"" +} +```",unknown,0,1 +xperience-10m-sample:qa:141,episode_qa,train,xperience-10m-sample,"{'start_frame': 2820, 'end_frame': 2839, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2820-2839"" +} +```",unknown,0,1 +xperience-10m-sample:qa:142,episode_qa,train,xperience-10m-sample,"{'start_frame': 2840, 'end_frame': 2859, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2840-2859"" +} +```",unknown,0,1 +xperience-10m-sample:qa:143,episode_qa,train,xperience-10m-sample,"{'start_frame': 2860, 'end_frame': 2879, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2860-2879"" +} +```",unknown,0,1 +xperience-10m-sample:qa:144,episode_qa,train,xperience-10m-sample,"{'start_frame': 2880, 'end_frame': 2899, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2880-2899"" +} +```",unknown,0,1 +xperience-10m-sample:qa:145,episode_qa,train,xperience-10m-sample,"{'start_frame': 2900, 'end_frame': 2919, 'num_frames': 20}",Grasp gooseneck kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""2900-2919"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:146,episode_qa,train,xperience-10m-sample,"{'start_frame': 2920, 'end_frame': 2939, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""2920-2939"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:147,episode_qa,train,xperience-10m-sample,"{'start_frame': 2940, 'end_frame': 2959, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2940-2959"" +} +```",unknown,0,1 +xperience-10m-sample:qa:148,episode_qa,train,xperience-10m-sample,"{'start_frame': 2960, 'end_frame': 2979, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:149,episode_qa,train,xperience-10m-sample,"{'start_frame': 2980, 'end_frame': 2999, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""2980-2999"" +} +```",unknown,0,1 +xperience-10m-sample:qa:150,episode_qa,train,xperience-10m-sample,"{'start_frame': 3000, 'end_frame': 3019, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3000-3019"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:151,episode_qa,train,xperience-10m-sample,"{'start_frame': 3020, 'end_frame': 3039, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3020-3039"" +} +```",unknown,0,1 +xperience-10m-sample:qa:152,episode_qa,train,xperience-10m-sample,"{'start_frame': 3040, 'end_frame': 3059, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3040-3059"" +} +```",unknown,0,1 +xperience-10m-sample:qa:153,episode_qa,train,xperience-10m-sample,"{'start_frame': 3060, 'end_frame': 3079, 'num_frames': 20}",Lift gooseneck kettle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3060-3079"" +} +```",unknown,0,1 +xperience-10m-sample:qa:170,episode_qa,train,xperience-10m-sample,"{'start_frame': 3400, 'end_frame': 3419, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3400-3419"" +} +```",unknown,0,1 +xperience-10m-sample:qa:171,episode_qa,train,xperience-10m-sample,"{'start_frame': 3420, 'end_frame': 3439, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3420-3439"" +} +```",unknown,0,1 +xperience-10m-sample:qa:172,episode_qa,train,xperience-10m-sample,"{'start_frame': 3440, 'end_frame': 3459, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3440-3459"" +} +```",unknown,0,1 +xperience-10m-sample:qa:173,episode_qa,train,xperience-10m-sample,"{'start_frame': 3460, 'end_frame': 3479, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3460-3479"" +} +```",unknown,0,1 +xperience-10m-sample:qa:174,episode_qa,train,xperience-10m-sample,"{'start_frame': 3480, 'end_frame': 3499, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3480-3499"" +} +```",unknown,0,1 +xperience-10m-sample:qa:175,episode_qa,train,xperience-10m-sample,"{'start_frame': 3500, 'end_frame': 3519, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3500-3519"" +} +```",unknown,0,1 +xperience-10m-sample:qa:176,episode_qa,train,xperience-10m-sample,"{'start_frame': 3520, 'end_frame': 3539, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3520-3539"" +} +```",unknown,0,1 +xperience-10m-sample:qa:177,episode_qa,train,xperience-10m-sample,"{'start_frame': 3540, 'end_frame': 3559, 'num_frames': 20}",Move kettle away,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3540-3559"" +} +```",unknown,0,1 +xperience-10m-sample:qa:178,episode_qa,train,xperience-10m-sample,"{'start_frame': 3560, 'end_frame': 3579, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3560-3579"" +} +```",unknown,0,1 +xperience-10m-sample:qa:179,episode_qa,train,xperience-10m-sample,"{'start_frame': 3580, 'end_frame': 3599, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3580-3599"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:180,episode_qa,train,xperience-10m-sample,"{'start_frame': 3600, 'end_frame': 3619, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:181,episode_qa,train,xperience-10m-sample,"{'start_frame': 3620, 'end_frame': 3639, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3620-3639"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:182,episode_qa,train,xperience-10m-sample,"{'start_frame': 3640, 'end_frame': 3659, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3640-3659"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:183,episode_qa,train,xperience-10m-sample,"{'start_frame': 3660, 'end_frame': 3679, 'num_frames': 20}",Place kettle on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3660-3679"" +} +```",unknown,0,1 +xperience-10m-sample:qa:184,episode_qa,train,xperience-10m-sample,"{'start_frame': 3680, 'end_frame': 3699, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3680-3699"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:185,episode_qa,train,xperience-10m-sample,"{'start_frame': 3700, 'end_frame': 3719, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3700-3719"" +} +```",unknown,0,1 +xperience-10m-sample:qa:186,episode_qa,train,xperience-10m-sample,"{'start_frame': 3720, 'end_frame': 3739, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3720-3739"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:187,episode_qa,train,xperience-10m-sample,"{'start_frame': 3740, 'end_frame': 3759, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3740-3759"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:188,episode_qa,train,xperience-10m-sample,"{'start_frame': 3760, 'end_frame': 3779, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3760-3779"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:189,episode_qa,train,xperience-10m-sample,"{'start_frame': 3780, 'end_frame': 3799, 'num_frames': 20}",Pick up white bottle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3780-3799"" +} +```",unknown,0,1 +xperience-10m-sample:qa:190,episode_qa,train,xperience-10m-sample,"{'start_frame': 3800, 'end_frame': 3819, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3800-3819"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:191,episode_qa,train,xperience-10m-sample,"{'start_frame': 3820, 'end_frame': 3839, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3820-3839"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:192,episode_qa,train,xperience-10m-sample,"{'start_frame': 3840, 'end_frame': 3859, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3840-3859"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:193,episode_qa,train,xperience-10m-sample,"{'start_frame': 3860, 'end_frame': 3879, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3860-3879"" +} +```",unknown,0,1 +xperience-10m-sample:qa:194,episode_qa,train,xperience-10m-sample,"{'start_frame': 3880, 'end_frame': 3899, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3880-3899"" +} +```",unknown,0,1 +xperience-10m-sample:qa:195,episode_qa,train,xperience-10m-sample,"{'start_frame': 3900, 'end_frame': 3919, 'num_frames': 20}",Pour liquid from white bottle,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3900-3919"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:196,episode_qa,train,xperience-10m-sample,"{'start_frame': 3920, 'end_frame': 3939, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:197,episode_qa,train,xperience-10m-sample,"{'start_frame': 3940, 'end_frame': 3959, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3940-3959"" +} +```",unknown,0,1 +xperience-10m-sample:qa:198,episode_qa,train,xperience-10m-sample,"{'start_frame': 3960, 'end_frame': 3979, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [""gooseneck kettle""], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""3960-3979"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:199,episode_qa,train,xperience-10m-sample,"{'start_frame': 3980, 'end_frame': 3999, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""3980-3999"" +} +```",unknown,0,1 +xperience-10m-sample:qa:200,episode_qa,train,xperience-10m-sample,"{'start_frame': 4000, 'end_frame': 4019, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [""gooseneck kettle""], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""4000-4019"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:201,episode_qa,train,xperience-10m-sample,"{'start_frame': 4020, 'end_frame': 4039, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4020-4039"" +} +```",unknown,0,1 +xperience-10m-sample:qa:202,episode_qa,train,xperience-10m-sample,"{'start_frame': 4040, 'end_frame': 4059, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""Pour milk into coffee"", + ""subtask"": ""Pour milk into coffee"", + ""objects"": [ + ""white bottle"", + ""coffee"" + ], + ""contact"": ""white bottle"", + ""transition"": ""Pouring milk into coffee"", + ""next_action"": ""Place white bottle on table"", + ""evidence_window"": ""4040-4059"" +} +```",Pour milk into coffee,0,1 +xperience-10m-sample:qa:203,episode_qa,train,xperience-10m-sample,"{'start_frame': 4060, 'end_frame': 4079, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4060-4079"" +} +```",unknown,0,1 +xperience-10m-sample:qa:204,episode_qa,train,xperience-10m-sample,"{'start_frame': 4080, 'end_frame': 4099, 'num_frames': 20}",Close bottle cap,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4080-4099"" +} +```",unknown,0,1 +xperience-10m-sample:qa:220,episode_qa,train,xperience-10m-sample,"{'start_frame': 4400, 'end_frame': 4419, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""unknown"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4400-4419"" +} +```",unknown,0,1 +xperience-10m-sample:qa:221,episode_qa,train,xperience-10m-sample,"{'start_frame': 4420, 'end_frame': 4439, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4420-4439"" +} +```",unknown,0,1 +xperience-10m-sample:qa:222,episode_qa,train,xperience-10m-sample,"{'start_frame': 4440, 'end_frame': 4459, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""4440-4459"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:223,episode_qa,train,xperience-10m-sample,"{'start_frame': 4460, 'end_frame': 4479, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""4460-4479"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:224,episode_qa,train,xperience-10m-sample,"{'start_frame': 4480, 'end_frame': 4499, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4480-4499"" +} +```",unknown,0,1 +xperience-10m-sample:qa:225,episode_qa,train,xperience-10m-sample,"{'start_frame': 4500, 'end_frame': 4519, 'num_frames': 20}",Place item on table,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4500-4519"" +} +```",unknown,0,1 +xperience-10m-sample:qa:226,episode_qa,train,xperience-10m-sample,"{'start_frame': 4520, 'end_frame': 4539, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""gooseneck kettle"", + ""white bottle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""unknown"" +} +```",unknown,0,1 +xperience-10m-sample:qa:227,episode_qa,train,xperience-10m-sample,"{'start_frame': 4540, 'end_frame': 4559, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4540-4559"" +} +```",unknown,0,1 +xperience-10m-sample:qa:228,episode_qa,train,xperience-10m-sample,"{'start_frame': 4560, 'end_frame': 4579, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4560-4579"" +} +```",unknown,0,1 +xperience-10m-sample:qa:229,episode_qa,train,xperience-10m-sample,"{'start_frame': 4580, 'end_frame': 4599, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4580-4599"" +} +```",unknown,0,1 +xperience-10m-sample:qa:230,episode_qa,train,xperience-10m-sample,"{'start_frame': 4600, 'end_frame': 4619, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4600-4619"" +} +```",unknown,0,1 +xperience-10m-sample:qa:231,episode_qa,train,xperience-10m-sample,"{'start_frame': 4620, 'end_frame': 4639, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4620-4639"" +} +```",unknown,0,1 +xperience-10m-sample:qa:232,episode_qa,train,xperience-10m-sample,"{'start_frame': 4640, 'end_frame': 4659, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4640-4659"" +} +```",unknown,0,1 +xperience-10m-sample:qa:233,episode_qa,train,xperience-10m-sample,"{'start_frame': 4660, 'end_frame': 4679, 'num_frames': 20}",Wait/Prepare for pouring,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""4660-4679"" +} +```",unknown,0,1 +xperience-10m-sample:qa:250,episode_qa,train,xperience-10m-sample,"{'start_frame': 5000, 'end_frame': 5019, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5000-5019"" +} +```",unknown,0,1 +xperience-10m-sample:qa:251,episode_qa,train,xperience-10m-sample,"{'start_frame': 5020, 'end_frame': 5039, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""Wait/Prepare for pouring"", + ""subtask"": ""Prepare for pouring"", + ""objects"": [ + ""gooseneck kettle"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""Pour coffee"", + ""evidence_window"": ""5020-5039"" +} +```",Wait/Prepare for pouring,0,1 +xperience-10m-sample:qa:252,episode_qa,train,xperience-10m-sample,"{'start_frame': 5040, 'end_frame': 5059, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5040-5059"" +} +```",unknown,0,1 +xperience-10m-sample:qa:253,episode_qa,train,xperience-10m-sample,"{'start_frame': 5060, 'end_frame': 5079, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5060-5079"" +} +```",unknown,0,1 +xperience-10m-sample:qa:254,episode_qa,train,xperience-10m-sample,"{'start_frame': 5080, 'end_frame': 5099, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5080-5099"" +} +```",unknown,0,1 +xperience-10m-sample:qa:255,episode_qa,train,xperience-10m-sample,"{'start_frame': 5100, 'end_frame': 5119, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5100-5119"" +} +```",unknown,0,1 +xperience-10m-sample:qa:256,episode_qa,train,xperience-10m-sample,"{'start_frame': 5120, 'end_frame': 5139, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5120-5139"" +} +```",unknown,0,1 +xperience-10m-sample:qa:257,episode_qa,train,xperience-10m-sample,"{'start_frame': 5140, 'end_frame': 5159, 'num_frames': 20}",Pour coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": ""unknown"", + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5140-5159"" +} +```",unknown,0,1 +xperience-10m-sample:qa:290,episode_qa,train,xperience-10m-sample,"{'start_frame': 5800, 'end_frame': 5819, 'num_frames': 20}",Pour milk into coffee,"```json +{ + ""action"": ""unknown"", + ""subtask"": ""unknown"", + ""objects"": [ + ""unknown"" + ], + ""contact"": ""unknown"", + ""transition"": ""unknown"", + ""next_action"": ""unknown"", + ""evidence_window"": ""5800-5819"" +} +```",unknown,0,1 diff --git a/results/omni_finetune/predictions_eval.jsonl b/results/omni_finetune/predictions_eval.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..7836e87e58b07e3717763d51b090da1457608672 --- /dev/null +++ b/results/omni_finetune/predictions_eval.jsonl @@ -0,0 +1,128 @@ +{"id": "xperience-10m-sample:qa:0", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 0, "end_frame": 19, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"0-19\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 0, "end_frame": 19}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "0-19"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:1", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 20, "end_frame": 39, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 20, "end_frame": 39}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:2", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 40, "end_frame": 59, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"40-59\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 40, "end_frame": 59}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "40-59"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:3", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 60, "end_frame": 79, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"60-79\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 60, "end_frame": 79}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "60-79"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:4", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 80, "end_frame": 99, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"80-99\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 80, "end_frame": 99}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "80-99"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:5", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 100, "end_frame": 119, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"100-119\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle", "coffee dripper", "scale", "bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 100, "end_frame": 119}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "100-119"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:6", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 120, "end_frame": 139, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"120-139\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Pick up kettle", "evidence_window": {"start_frame": 120, "end_frame": 139}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "120-139"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:7", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 140, "end_frame": 159, "num_frames": 20}, "true_label": "Pick up kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"140-159\"\n}\n```", "true_json": {"action": "Pick up kettle", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 140, "end_frame": 159}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "140-159"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:8", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 160, "end_frame": 179, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"160-179\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 160, "end_frame": 179}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "160-179"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:9", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 180, "end_frame": 199, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"180-199\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 180, "end_frame": 199}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "180-199"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:10", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 200, "end_frame": 219, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"200-219\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 200, "end_frame": 219}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "200-219"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:11", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 220, "end_frame": 239, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"220-239\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 220, "end_frame": 239}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "220-239"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:12", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 240, "end_frame": 259, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"240-259\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 240, "end_frame": 259}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "240-259"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:13", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 260, "end_frame": 279, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"260-279\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 260, "end_frame": 279}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "260-279"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:14", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 280, "end_frame": 299, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"280-299\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 280, "end_frame": 299}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "280-299"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:15", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 300, "end_frame": 319, "num_frames": 20}, "true_label": "Position kettle to pour", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Position kettle to pour", "subtask": "Pick up and position kettle", "objects": ["kettle"], "contact": "yes", "transition": "no", "next_action": "Position kettle to pour", "evidence_window": {"start_frame": 300, "end_frame": 319}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:40", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 800, "end_frame": 819, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"800-819\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 800, "end_frame": 819}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "800-819"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:41", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 820, "end_frame": 839, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"820-839\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 820, "end_frame": 839}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "820-839"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:42", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 840, "end_frame": 859, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 840, "end_frame": 859}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:43", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 860, "end_frame": 879, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 860, "end_frame": 879}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:44", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 880, "end_frame": 899, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"880-899\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 880, "end_frame": 899}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "880-899"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:45", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 900, "end_frame": 919, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"900-919\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 900, "end_frame": 919}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "900-919"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:46", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 920, "end_frame": 939, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"920-939\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 920, "end_frame": 939}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "920-939"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:47", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 940, "end_frame": 959, "num_frames": 20}, "true_label": "Move kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"940-959\"\n}\n```", "true_json": {"action": "Move kettle", "subtask": "Move kettle", "objects": ["kettle", "dripper", "scale", "coffee filter", "table"], "contact": "yes", "transition": "no", "next_action": "Move kettle", "evidence_window": {"start_frame": 940, "end_frame": 959}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "940-959"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:50", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1000, "end_frame": 1019, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1000-1019\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1000, "end_frame": 1019}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1000-1019"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:51", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1020, "end_frame": 1039, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1020-1039\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1020, "end_frame": 1039}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1020-1039"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:52", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1040, "end_frame": 1059, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1040, "end_frame": 1059}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:53", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1060, "end_frame": 1079, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1060-1079\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1060, "end_frame": 1079}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1060-1079"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:54", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1080, "end_frame": 1099, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1080-1099\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1080, "end_frame": 1099}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1080-1099"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:55", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1100, "end_frame": 1119, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1100-1119\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1100, "end_frame": 1119}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1100-1119"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:56", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1120, "end_frame": 1139, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1120-1139\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1120, "end_frame": 1139}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1120-1139"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:57", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1140, "end_frame": 1159, "num_frames": 20}, "true_label": "Hold coffee carafe", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1140-1159\"\n}\n```", "true_json": {"action": "Hold coffee carafe", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "coffee scale", "coffee dripper", "white mug", "bottle"], "contact": "yes", "transition": "no", "next_action": "Hold coffee carafe", "evidence_window": {"start_frame": 1140, "end_frame": 1159}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1140-1159"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:62", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1240, "end_frame": 1259, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\"gooseneck kettle\"],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1240-1259\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1240, "end_frame": 1259}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1240-1259"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:63", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1260, "end_frame": 1279, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1260-1279\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1260, "end_frame": 1279}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1260-1279"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:64", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1280, "end_frame": 1299, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1280-1299\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1280, "end_frame": 1299}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1280-1299"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:65", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1300, "end_frame": 1319, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1300-1319\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1300, "end_frame": 1319}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1300-1319"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:66", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1320, "end_frame": 1339, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1320, "end_frame": 1339}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:67", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1340, "end_frame": 1359, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1340-1359\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1340, "end_frame": 1359}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1340-1359"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:68", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1360, "end_frame": 1379, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1360-1379\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1360, "end_frame": 1379}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1360-1379"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:69", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1380, "end_frame": 1399, "num_frames": 20}, "true_label": "Grasp coffee scoop", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1380-1399\"\n}\n```", "true_json": {"action": "Grasp coffee scoop", "subtask": "Prepare coffee equipment and scoop grounds", "objects": ["glass carafe", "wooden scoop"], "contact": "yes", "transition": "no", "next_action": "Grasp coffee scoop", "evidence_window": {"start_frame": 1380, "end_frame": 1399}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1380-1399"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:78", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1560, "end_frame": 1579, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1560-1579\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee jar", "coffee scoop"], "contact": "yes", "transition": "yes", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1560, "end_frame": 1579}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1560-1579"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:79", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1580, "end_frame": 1599, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1580-1599\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee jar", "coffee scoop"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1580, "end_frame": 1599}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1580-1599"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:80", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1600, "end_frame": 1619, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1600-1619\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1600, "end_frame": 1619}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1600-1619"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:81", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1620, "end_frame": 1639, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1620-1639\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1620, "end_frame": 1639}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1620-1639"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:82", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1640, "end_frame": 1659, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1640-1659\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1640, "end_frame": 1659}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1640-1659"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:83", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1660, "end_frame": 1679, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1660-1679\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1660, "end_frame": 1679}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1660-1679"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:84", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1680, "end_frame": 1699, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"1680-1699\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1680, "end_frame": 1699}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "1680-1699"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:85", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 1700, "end_frame": 1719, "num_frames": 20}, "true_label": "Transfer coffee to dripper", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"1700-1719\"\n}\n```", "true_json": {"action": "Transfer coffee to dripper", "subtask": "Transfer coffee grounds to dripper", "objects": ["coffee scoop", "dripper"], "contact": "yes", "transition": "no", "next_action": "Transfer coffee to dripper", "evidence_window": {"start_frame": 1700, "end_frame": 1719}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "1700-1719"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:100", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2000, "end_frame": 2019, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2000-2019\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2000, "end_frame": 2019}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2000-2019"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:101", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2020, "end_frame": 2039, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2020-2039\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2020, "end_frame": 2039}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2020-2039"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:102", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2040, "end_frame": 2059, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2040-2059\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2040, "end_frame": 2059}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2040-2059"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:103", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2060, "end_frame": 2079, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2060-2079\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2060, "end_frame": 2079}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2060-2079"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:104", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2080, "end_frame": 2099, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2080-2099\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2080, "end_frame": 2099}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2080-2099"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:105", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2100, "end_frame": 2119, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2100-2119\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2100, "end_frame": 2119}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2100-2119"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:106", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2120, "end_frame": 2139, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"unknown\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2120-2139\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2120, "end_frame": 2139}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["unknown"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2120-2139"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:107", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2140, "end_frame": 2159, "num_frames": 20}, "true_label": "Hold gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2140-2159\"\n}\n```", "true_json": {"action": "Hold gooseneck kettle", "subtask": "Handle gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "water bottle", "coffee mug"], "contact": "yes", "transition": "no", "next_action": "Hold gooseneck kettle", "evidence_window": {"start_frame": 2140, "end_frame": 2159}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2140-2159"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:140", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2800, "end_frame": 2819, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2800-2819\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2800, "end_frame": 2819}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2800-2819"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:141", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2820, "end_frame": 2839, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2820-2839\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2820, "end_frame": 2839}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2820-2839"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:142", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2840, "end_frame": 2859, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2840-2859\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2840, "end_frame": 2859}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2840-2859"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:143", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2860, "end_frame": 2879, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2860-2879\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2860, "end_frame": 2879}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2860-2879"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:144", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2880, "end_frame": 2899, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2880-2899\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Grasp gooseneck kettle", "evidence_window": {"start_frame": 2880, "end_frame": 2899}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2880-2899"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:145", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2900, "end_frame": 2919, "num_frames": 20}, "true_label": "Grasp gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"2900-2919\"\n}\n```", "true_json": {"action": "Grasp gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2900, "end_frame": 2919}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "2900-2919"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:146", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2920, "end_frame": 2939, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"2920-2939\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2920, "end_frame": 2939}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "2920-2939"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:147", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2940, "end_frame": 2959, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2940-2959\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "water bottle", "white cup"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2940, "end_frame": 2959}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2940-2959"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:148", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2960, "end_frame": 2979, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2960, "end_frame": 2979}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:149", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 2980, "end_frame": 2999, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"2980-2999\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 2980, "end_frame": 2999}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "2980-2999"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:150", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3000, "end_frame": 3019, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3000-3019\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3000, "end_frame": 3019}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3000-3019"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:151", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3020, "end_frame": 3039, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3020-3039\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3020, "end_frame": 3039}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3020-3039"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:152", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3040, "end_frame": 3059, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3040-3059\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3040, "end_frame": 3059}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3040-3059"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:153", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3060, "end_frame": 3079, "num_frames": 20}, "true_label": "Lift gooseneck kettle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3060-3079\"\n}\n```", "true_json": {"action": "Lift gooseneck kettle", "subtask": "Lift gooseneck kettle", "objects": ["gooseneck kettle", "coffee dripper", "digital scale", "white mug"], "contact": "yes", "transition": "no", "next_action": "Lift gooseneck kettle", "evidence_window": {"start_frame": 3060, "end_frame": 3079}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3060-3079"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:170", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3400, "end_frame": 3419, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3400-3419\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3400, "end_frame": 3419}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3400-3419"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:171", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3420, "end_frame": 3439, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3420-3439\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3420, "end_frame": 3439}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3420-3439"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:172", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3440, "end_frame": 3459, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3440-3459\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3440, "end_frame": 3459}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3440-3459"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:173", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3460, "end_frame": 3479, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3460-3479\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3460, "end_frame": 3479}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3460-3479"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:174", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3480, "end_frame": 3499, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3480-3499\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3480, "end_frame": 3499}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3480-3499"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:175", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3500, "end_frame": 3519, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3500-3519\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3500, "end_frame": 3519}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3500-3519"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:176", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3520, "end_frame": 3539, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3520-3539\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Move kettle away", "evidence_window": {"start_frame": 3520, "end_frame": 3539}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3520-3539"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:177", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3540, "end_frame": 3559, "num_frames": 20}, "true_label": "Move kettle away", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3540-3559\"\n}\n```", "true_json": {"action": "Move kettle away", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3540, "end_frame": 3559}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3540-3559"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:178", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3560, "end_frame": 3579, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3560-3579\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3560, "end_frame": 3579}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3560-3579"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:179", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3580, "end_frame": 3599, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3580-3599\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3580, "end_frame": 3599}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3580-3599"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:180", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3600, "end_frame": 3619, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3600, "end_frame": 3619}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:181", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3620, "end_frame": 3639, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3620-3639\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3620, "end_frame": 3639}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3620-3639"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:182", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3640, "end_frame": 3659, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3640-3659\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Place kettle on table", "evidence_window": {"start_frame": 3640, "end_frame": 3659}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3640-3659"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:183", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3660, "end_frame": 3679, "num_frames": 20}, "true_label": "Place kettle on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3660-3679\"\n}\n```", "true_json": {"action": "Place kettle on table", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3660, "end_frame": 3679}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3660-3679"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:184", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3680, "end_frame": 3699, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3680-3699\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3680, "end_frame": 3699}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3680-3699"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:185", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3700, "end_frame": 3719, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3700-3719\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3700, "end_frame": 3719}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3700-3719"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:186", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3720, "end_frame": 3739, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3720-3739\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3720, "end_frame": 3739}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3720-3739"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:187", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3740, "end_frame": 3759, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3740-3759\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3740, "end_frame": 3759}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3740-3759"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:188", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3760, "end_frame": 3779, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3760-3779\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pick up white bottle", "evidence_window": {"start_frame": 3760, "end_frame": 3779}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3760-3779"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:189", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3780, "end_frame": 3799, "num_frames": 20}, "true_label": "Pick up white bottle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3780-3799\"\n}\n```", "true_json": {"action": "Pick up white bottle", "subtask": "Set down kettle and retrieve white bottle", "objects": ["gooseneck kettle", "coffee dripper", "scale", "white bottle"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3780, "end_frame": 3799}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3780-3799"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:190", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3800, "end_frame": 3819, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3800-3819\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3800, "end_frame": 3819}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3800-3819"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:191", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3820, "end_frame": 3839, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3820-3839\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3820, "end_frame": 3839}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3820-3839"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:192", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3840, "end_frame": 3859, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3840-3859\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3840, "end_frame": 3859}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3840-3859"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:193", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3860, "end_frame": 3879, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3860-3879\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3860, "end_frame": 3879}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3860-3879"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:194", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3880, "end_frame": 3899, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3880-3899\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour liquid from white bottle", "evidence_window": {"start_frame": 3880, "end_frame": 3899}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3880-3899"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:195", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3900, "end_frame": 3919, "num_frames": 20}, "true_label": "Pour liquid from white bottle", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3900-3919\"\n}\n```", "true_json": {"action": "Pour liquid from white bottle", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3900, "end_frame": 3919}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3900-3919"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:196", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3920, "end_frame": 3939, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3920, "end_frame": 3939}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:197", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3940, "end_frame": 3959, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3940-3959\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3940, "end_frame": 3959}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3940-3959"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:198", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3960, "end_frame": 3979, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\"gooseneck kettle\"],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"3960-3979\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3960, "end_frame": 3979}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "3960-3979"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:199", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 3980, "end_frame": 3999, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"3980-3999\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 3980, "end_frame": 3999}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "3980-3999"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:200", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4000, "end_frame": 4019, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\"gooseneck kettle\"],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"4000-4019\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4000, "end_frame": 4019}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "4000-4019"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:201", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4020, "end_frame": 4039, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4020-4039\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4020, "end_frame": 4039}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4020-4039"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:202", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4040, "end_frame": 4059, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"Pour milk into coffee\",\n \"subtask\": \"Pour milk into coffee\",\n \"objects\": [\n \"white bottle\",\n \"coffee\"\n ],\n \"contact\": \"white bottle\",\n \"transition\": \"Pouring milk into coffee\",\n \"next_action\": \"Place white bottle on table\",\n \"evidence_window\": \"4040-4059\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4040, "end_frame": 4059}}, "pred_json": {"action": "Pour milk into coffee", "subtask": "Pour milk into coffee", "objects": ["white bottle", "coffee"], "contact": "white bottle", "transition": "Pouring milk into coffee", "next_action": "Place white bottle on table", "evidence_window": "4040-4059"}, "predicted_label": "Pour milk into coffee", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:203", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4060, "end_frame": 4079, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4060-4079\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4060, "end_frame": 4079}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4060-4079"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:204", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4080, "end_frame": 4099, "num_frames": 20}, "true_label": "Close bottle cap", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4080-4099\"\n}\n```", "true_json": {"action": "Close bottle cap", "subtask": "Pour and close white bottle", "objects": ["white bottle", "coffee dripper", "scale", "mug"], "contact": "yes", "transition": "no", "next_action": "Close bottle cap", "evidence_window": {"start_frame": 4080, "end_frame": 4099}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4080-4099"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:220", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4400, "end_frame": 4419, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"unknown\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4400-4419\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4400, "end_frame": 4419}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["unknown"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4400-4419"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:221", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4420, "end_frame": 4439, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4420-4439\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4420, "end_frame": 4439}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4420-4439"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:222", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4440, "end_frame": 4459, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"4440-4459\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4440, "end_frame": 4459}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "4440-4459"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:223", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4460, "end_frame": 4479, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"4460-4479\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4460, "end_frame": 4479}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "4460-4479"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:224", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4480, "end_frame": 4499, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4480-4499\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Place item on table", "evidence_window": {"start_frame": 4480, "end_frame": 4499}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4480-4499"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:225", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4500, "end_frame": 4519, "num_frames": 20}, "true_label": "Place item on table", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4500-4519\"\n}\n```", "true_json": {"action": "Place item on table", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4500, "end_frame": 4519}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4500-4519"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:226", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4520, "end_frame": 4539, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"gooseneck kettle\",\n \"white bottle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"unknown\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4520, "end_frame": 4539}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["gooseneck kettle", "white bottle"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "unknown"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:227", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4540, "end_frame": 4559, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4540-4559\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4540, "end_frame": 4559}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4540-4559"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:228", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4560, "end_frame": 4579, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4560-4579\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4560, "end_frame": 4579}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4560-4579"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:229", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4580, "end_frame": 4599, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4580-4599\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4580, "end_frame": 4599}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4580-4599"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:230", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4600, "end_frame": 4619, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4600-4619\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4600, "end_frame": 4619}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4600-4619"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:231", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4620, "end_frame": 4639, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4620-4639\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4620, "end_frame": 4639}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4620-4639"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:232", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4640, "end_frame": 4659, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4640-4659\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4640, "end_frame": 4659}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4640-4659"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:233", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 4660, "end_frame": 4679, "num_frames": 20}, "true_label": "Wait/Prepare for pouring", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"4660-4679\"\n}\n```", "true_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["coffee equipment", "small bottle", "white mug", "weighing scale"], "contact": "yes", "transition": "no", "next_action": "Wait/Prepare for pouring", "evidence_window": {"start_frame": 4660, "end_frame": 4679}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "4660-4679"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:250", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5000, "end_frame": 5019, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5000-5019\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5000, "end_frame": 5019}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5000-5019"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:251", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5020, "end_frame": 5039, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"Wait/Prepare for pouring\",\n \"subtask\": \"Prepare for pouring\",\n \"objects\": [\n \"gooseneck kettle\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"Pour coffee\",\n \"evidence_window\": \"5020-5039\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5020, "end_frame": 5039}}, "pred_json": {"action": "Wait/Prepare for pouring", "subtask": "Prepare for pouring", "objects": ["gooseneck kettle"], "contact": "unknown", "transition": "unknown", "next_action": "Pour coffee", "evidence_window": "5020-5039"}, "predicted_label": "Wait/Prepare for pouring", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:252", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5040, "end_frame": 5059, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5040-5059\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5040, "end_frame": 5059}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5040-5059"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:253", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5060, "end_frame": 5079, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5060-5079\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5060, "end_frame": 5079}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5060-5079"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:254", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5080, "end_frame": 5099, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5080-5099\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5080, "end_frame": 5099}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5080-5099"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:255", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5100, "end_frame": 5119, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5100-5119\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5100, "end_frame": 5119}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5100-5119"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:256", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5120, "end_frame": 5139, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5120-5139\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5120, "end_frame": 5139}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": [], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5120-5139"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:257", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5140, "end_frame": 5159, "num_frames": 20}, "true_label": "Pour coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": \"unknown\",\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5140-5159\"\n}\n```", "true_json": {"action": "Pour coffee", "subtask": "Pour coffee", "objects": ["carafe", "coffee mug", "scale"], "contact": "yes", "transition": "no", "next_action": "Pour coffee", "evidence_window": {"start_frame": 5140, "end_frame": 5159}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": "unknown", "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5140-5159"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} +{"id": "xperience-10m-sample:qa:290", "target": "episode_qa", "split": "train", "episode_id": "xperience-10m-sample", "center_window": {"start_frame": 5800, "end_frame": 5819, "num_frames": 20}, "true_label": "Pour milk into coffee", "raw_prediction": "```json\n{\n \"action\": \"unknown\",\n \"subtask\": \"unknown\",\n \"objects\": [\n \"unknown\"\n ],\n \"contact\": \"unknown\",\n \"transition\": \"unknown\",\n \"next_action\": \"unknown\",\n \"evidence_window\": \"5800-5819\"\n}\n```", "true_json": {"action": "Pour milk into coffee", "subtask": "Pour milk into coffee", "objects": ["coffee mug", "stainless steel milk pitcher", "table", "milk bottle", "digital scale"], "contact": "yes", "transition": "no", "next_action": "Pour milk into coffee", "evidence_window": {"start_frame": 5800, "end_frame": 5819}}, "pred_json": {"action": "unknown", "subtask": "unknown", "objects": ["unknown"], "contact": "unknown", "transition": "unknown", "next_action": "unknown", "evidence_window": "5800-5819"}, "predicted_label": "unknown", "correct": 0, "true_label_seen_in_train": 1} diff --git a/results/omni_finetune/progress.jsonl b/results/omni_finetune/progress.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..27f5a2158f802df7a329896a893aa73664574f84 --- /dev/null +++ b/results/omni_finetune/progress.jsonl @@ -0,0 +1,19 @@ +{"event": "start", "run_id": "xperience10m_qwen3_omni_32ep_lora", "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", "num_processes": 8, "num_train_samples": 128, "num_val_samples": 0, "rank_samples_per_epoch": 16, "epochs": 1, "timestamp": 1780145767.3727567} +{"event": "train_step", "epoch": 1, "global_step": 1, "rank0_seen": 1, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.024569511413574, "timestamp": 1780145769.5535998} +{"event": "train_step", "epoch": 1, "global_step": 2, "rank0_seen": 2, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.932429313659668, "timestamp": 1780145770.3543525} +{"event": "train_step", "epoch": 1, "global_step": 3, "rank0_seen": 3, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.078858375549316, "timestamp": 1780145771.1473835} +{"event": "train_step", "epoch": 1, "global_step": 4, "rank0_seen": 4, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.395858764648438, "timestamp": 1780145771.93893} +{"event": "train_step", "epoch": 1, "global_step": 5, "rank0_seen": 5, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.80813217163086, "timestamp": 1780145772.7313085} +{"event": "train_step", "epoch": 1, "global_step": 6, "rank0_seen": 6, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.644664764404297, "timestamp": 1780145773.528463} +{"event": "train_step", "epoch": 1, "global_step": 7, "rank0_seen": 7, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.009082794189453, "timestamp": 1780145774.3242047} +{"event": "train_step", "epoch": 1, "global_step": 8, "rank0_seen": 8, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.84404468536377, "timestamp": 1780145775.4113226} +{"event": "train_step", "epoch": 1, "global_step": 9, "rank0_seen": 9, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.985627174377441, "timestamp": 1780145776.5403607} +{"event": "train_step", "epoch": 1, "global_step": 10, "rank0_seen": 10, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.62201976776123, "timestamp": 1780145777.339885} +{"event": "train_step", "epoch": 1, "global_step": 11, "rank0_seen": 11, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.205351829528809, "timestamp": 1780145778.1362987} +{"event": "train_step", "epoch": 1, "global_step": 12, "rank0_seen": 12, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.856507301330566, "timestamp": 1780145778.9328642} +{"event": "train_step", "epoch": 1, "global_step": 13, "rank0_seen": 13, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.044717788696289, "timestamp": 1780145779.7260861} +{"event": "train_step", "epoch": 1, "global_step": 14, "rank0_seen": 14, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.76926326751709, "timestamp": 1780145780.5313745} +{"event": "train_step", "epoch": 1, "global_step": 15, "rank0_seen": 15, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.089532852172852, "timestamp": 1780145781.3197153} +{"event": "train_step", "epoch": 1, "global_step": 16, "rank0_seen": 16, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.671160697937012, "timestamp": 1780145782.1418438} +{"event": "epoch_end", "epoch": 1, "train_loss": 10.936363816261292, "val_loss": null, "global_step": 16, "timestamp": 1780145782.1423175} +{"event": "complete", "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", "timestamp": 1780145782.743853} diff --git a/results/omni_finetune/progress_lora.jsonl b/results/omni_finetune/progress_lora.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..27f5a2158f802df7a329896a893aa73664574f84 --- /dev/null +++ b/results/omni_finetune/progress_lora.jsonl @@ -0,0 +1,19 @@ +{"event": "start", "run_id": "xperience10m_qwen3_omni_32ep_lora", "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", "num_processes": 8, "num_train_samples": 128, "num_val_samples": 0, "rank_samples_per_epoch": 16, "epochs": 1, "timestamp": 1780145767.3727567} +{"event": "train_step", "epoch": 1, "global_step": 1, "rank0_seen": 1, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.024569511413574, "timestamp": 1780145769.5535998} +{"event": "train_step", "epoch": 1, "global_step": 2, "rank0_seen": 2, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.932429313659668, "timestamp": 1780145770.3543525} +{"event": "train_step", "epoch": 1, "global_step": 3, "rank0_seen": 3, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.078858375549316, "timestamp": 1780145771.1473835} +{"event": "train_step", "epoch": 1, "global_step": 4, "rank0_seen": 4, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.395858764648438, "timestamp": 1780145771.93893} +{"event": "train_step", "epoch": 1, "global_step": 5, "rank0_seen": 5, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.80813217163086, "timestamp": 1780145772.7313085} +{"event": "train_step", "epoch": 1, "global_step": 6, "rank0_seen": 6, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.644664764404297, "timestamp": 1780145773.528463} +{"event": "train_step", "epoch": 1, "global_step": 7, "rank0_seen": 7, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.009082794189453, "timestamp": 1780145774.3242047} +{"event": "train_step", "epoch": 1, "global_step": 8, "rank0_seen": 8, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.84404468536377, "timestamp": 1780145775.4113226} +{"event": "train_step", "epoch": 1, "global_step": 9, "rank0_seen": 9, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.985627174377441, "timestamp": 1780145776.5403607} +{"event": "train_step", "epoch": 1, "global_step": 10, "rank0_seen": 10, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.62201976776123, "timestamp": 1780145777.339885} +{"event": "train_step", "epoch": 1, "global_step": 11, "rank0_seen": 11, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.205351829528809, "timestamp": 1780145778.1362987} +{"event": "train_step", "epoch": 1, "global_step": 12, "rank0_seen": 12, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.856507301330566, "timestamp": 1780145778.9328642} +{"event": "train_step", "epoch": 1, "global_step": 13, "rank0_seen": 13, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.044717788696289, "timestamp": 1780145779.7260861} +{"event": "train_step", "epoch": 1, "global_step": 14, "rank0_seen": 14, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.76926326751709, "timestamp": 1780145780.5313745} +{"event": "train_step", "epoch": 1, "global_step": 15, "rank0_seen": 15, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 11.089532852172852, "timestamp": 1780145781.3197153} +{"event": "train_step", "epoch": 1, "global_step": 16, "rank0_seen": 16, "rank0_samples_per_epoch": 16, "rank0_batch_loss": 10.671160697937012, "timestamp": 1780145782.1418438} +{"event": "epoch_end", "epoch": 1, "train_loss": 10.936363816261292, "val_loss": null, "global_step": 16, "timestamp": 1780145782.1423175} +{"event": "complete", "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", "timestamp": 1780145782.743853} diff --git a/results/omni_finetune/source_discovery.json b/results/omni_finetune/source_discovery.json new file mode 100644 index 0000000000000000000000000000000000000000..650a7bd87ae1a328c403fd00c28bdd8aef7db02e --- /dev/null +++ b/results/omni_finetune/source_discovery.json @@ -0,0 +1,57 @@ +{ + "target_episodes": 32, + "workspace": "/path/to/ropedia_workspace/ropedia-episode-task-suite", + "data_root": "/path/to/ropedia_workspace/modelscope_data", + "ready_for_32_episode_pilot": false, + "selected_source": "none", + "local": { + "source": "local", + "num_episodes": 1, + "num_degraded_valid_episodes": 1, + "num_complete_episodes": 0, + "errors": [], + "episodes": [ + { + "source": "local", + "episode_id": "xperience-10m-sample", + "episode_path": "/path/to/ropedia_workspace/modelscope_data/xperience-10m-sample", + "has_annotation": true, + "has_fisheye_cam0": true, + "has_all_videos": false, + "has_any_video": true, + "missing_views": [ + "fisheye_cam1.mp4", + "fisheye_cam2.mp4", + "fisheye_cam3.mp4", + "stereo_left.mp4", + "stereo_right.mp4" + ], + "is_degraded_valid": true, + "is_complete": false + } + ], + "data_root": "/path/to/ropedia_workspace/modelscope_data" + }, + "modelscope": { + "source": "modelscope", + "num_episodes": 0, + "num_degraded_valid_episodes": 0, + "num_complete_episodes": 0, + "errors": [], + "episodes": [] + }, + "huggingface": { + "source": "huggingface", + "num_episodes": 0, + "num_degraded_valid_episodes": 0, + "num_complete_episodes": 0, + "errors": [], + "episodes": [] + }, + "blockers": [ + "Not enough degraded-valid episodes for a 32-episode pilot. Need 32, local has 1.", + "Current training host path remains one-episode proof-of-stack only.", + "ModelScope probe unavailable or reported no matching episode files.", + "Hugging Face probe unavailable or reported no matching episode files." + ] +} \ No newline at end of file diff --git a/results/omni_finetune/training_metadata.json b/results/omni_finetune/training_metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..50814ee383d9cf91fe31fbdd5e76a2cdacff9215 --- /dev/null +++ b/results/omni_finetune/training_metadata.json @@ -0,0 +1,32 @@ +{ + "run_id": "xperience10m_qwen3_omni_32ep_lora", + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "num_processes": 8, + "num_train_samples": 128, + "num_val_samples": 0, + "history": [ + { + "epoch": 1, + "train_loss": 10.936363816261292, + "val_loss": null, + "global_step": 16 + } + ], + "lora": { + "r": 16, + "alpha": 32, + "dropout": 0.05, + "target_modules": [ + "q_proj", + "k_proj", + "v_proj", + "o_proj", + "gate_proj", + "up_proj", + "down_proj" + ] + }, + "use_audio_in_video": false +} \ No newline at end of file diff --git a/results/omni_finetune/training_metadata_lora.json b/results/omni_finetune/training_metadata_lora.json new file mode 100644 index 0000000000000000000000000000000000000000..50814ee383d9cf91fe31fbdd5e76a2cdacff9215 --- /dev/null +++ b/results/omni_finetune/training_metadata_lora.json @@ -0,0 +1,32 @@ +{ + "run_id": "xperience10m_qwen3_omni_32ep_lora", + "model_id": "/path/to/ropedia_workspace/modelscope_models/Qwen__Qwen3-Omni-30B-A3B-Instruct", + "dataset_jsonl": "results/omni_finetune/xperience10m_qwen3_omni_32ep_dataset/dataset.jsonl", + "checkpoint_dir": "/path/to/ropedia_workspace/ropedia-episode-task-suite/checkpoints/xperience10m_qwen3_omni_32ep_lora/adapter_lora", + "num_processes": 8, + "num_train_samples": 128, + "num_val_samples": 0, + "history": [ + { + "epoch": 1, + "train_loss": 10.936363816261292, + "val_loss": null, + "global_step": 16 + } + ], + "lora": { + "r": 16, + "alpha": 32, + "dropout": 0.05, + "target_modules": [ + "q_proj", + "k_proj", + "v_proj", + "o_proj", + "gate_proj", + "up_proj", + "down_proj" + ] + }, + "use_audio_in_video": false +} \ No newline at end of file diff --git a/scripts/build_artifact_index.py b/scripts/build_artifact_index.py index 81f0885e6c62c015f38f83060eb7d48816ad8339..2a071cf5df4d709fede86b0450bd07f92adee597 100644 --- a/scripts/build_artifact_index.py +++ b/scripts/build_artifact_index.py @@ -39,7 +39,7 @@ ARTIFACTS = [ "path": "EVIDENCE_CONTRACT.md", "kind": "claim_boundary", "surface": "repo", - "proves": "Defines what is verified, what is smoke-only, and what must not be inferred.", + "proves": "Defines what is verified, what is readiness-only, and what must not be inferred.", }, { "id": "reviewer_packet", @@ -391,12 +391,12 @@ ARTIFACTS = [ "proves": "Documents why no 32-episode Qwen3-Omni result is claimed yet.", }, { - "id": "a100_relay_status", - "title": "A100 relay status", - "path": "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "id": "multi_episode_access_status", + "title": "Multi-episode access status", + "path": "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", "kind": "scaleup_status", "surface": "repo_hf", - "proves": "Documents the pending A100-to-H20 data relay and 32-session pilot selection.", + "proves": "Documents the public multi-episode access boundary and 32-episode pilot selection without exposing private infrastructure details.", }, { "id": "citation", diff --git a/scripts/build_brand_assets.py b/scripts/build_brand_assets.py index dacea4296fd3f4aa60c5aa80423b230922e7a30f..60c4b447d3007a08dc4a9c53f4e66a3c32c0c8f0 100644 --- a/scripts/build_brand_assets.py +++ b/scripts/build_brand_assets.py @@ -215,7 +215,7 @@ def main() -> int: "source": { "path": SOURCE_MARK.relative_to(ROOT).as_posix(), "kind": "ChatGPT-image-generated logo mark with chroma-key background removed locally", - "prompt_summary": "X-shaped multimodal camera mark with Ropedia-inspired near-black, lime, cyan, trajectory, and point-cloud styling.", + "prompt_summary": "X-shaped multimodal camera mark with near-black, lime, cyan, trajectory, and point-cloud styling.", }, "assets": [ image_record("logo_mark", SOURCE_MARK, "Transparent source logo mark."), diff --git a/scripts/build_evaluation_protocol.py b/scripts/build_evaluation_protocol.py index 88b342d725649d30c243d9cd87474d94704bc644..3aae76a520007a58fa9516db238497e09d56e822 100644 --- a/scripts/build_evaluation_protocol.py +++ b/scripts/build_evaluation_protocol.py @@ -226,7 +226,7 @@ def build_payload() -> dict: "unsupported_interpretations": [ "Do not infer cross-episode generalization from this single public sample.", "Do not treat feature-vector reconstruction as pixel depth, mesh, NeRF, or Gaussian reconstruction.", - "Do not treat Qwen3-Omni smoke artifacts as a real 32-episode fine-tune.", + "Do not treat Qwen3-Omni readiness artifacts as a real 32-episode fine-tune.", "Do not infer audio-visual learning from the current baseline vector because audio is not featurized.", ], "scale_up_gate": { @@ -239,7 +239,7 @@ def build_payload() -> dict: "current_status": "prepared but data-gated", "evidence": [ "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", ], }, } @@ -350,7 +350,7 @@ def render_markdown(payload: dict) -> str: "", "Current status: prepared but data-gated. Read", "`results/omni_finetune/DATA_BLOCKER_REPORT.md` and", - "`results/omni_finetune/A100_HF_RELAY_STATUS.md` before interpreting any", + "`results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md` before interpreting any", "Qwen3-Omni artifact.", "", "## Machine-Readable Copy", diff --git a/scripts/build_quality_gates.py b/scripts/build_quality_gates.py index 728a417f0e1fbb099790fccd44ba589bb397deba..4db0b61580e9e27b7a9577f5d8a3edb19ca96346 100644 --- a/scripts/build_quality_gates.py +++ b/scripts/build_quality_gates.py @@ -24,8 +24,8 @@ GATES = [ "title": "Scope claims guard", "command": "python scripts/validate_scope_claims.py", "report": "docs/data/scope_claims_audit.json", - "blocks_if": "Historical 32ep smoke/provenance strings are presented as real 32-episode metrics.", - "proves": "The public narrative does not overclaim the Qwen3-Omni smoke artifacts.", + "blocks_if": "Historical 32ep readiness/provenance strings are presented as real 32-episode metrics.", + "proves": "The public narrative does not overclaim the Qwen3-Omni readiness artifacts.", }, { "id": "source_alignment", @@ -115,8 +115,8 @@ POST_PUBLISH_CHECKS = [ "required_result": "latest pages-build-deployment run succeeds", }, { - "id": "rendered_browser_smoke", - "title": "Rendered browser smoke", + "id": "rendered_browser_check", + "title": "Rendered browser check", "evidence": "Browser/Playwright page identity, nonblank render, console health, and one local interaction", "required_result": "no relevant console warnings/errors and target links work", }, diff --git a/scripts/generate_visualizations.py b/scripts/generate_visualizations.py index 3656ab13e095ad45d55064e0a86eba6c2a0dda93..a3c57f233bdd7c3a9e111f4ef937c3918f648c7a 100644 --- a/scripts/generate_visualizations.py +++ b/scripts/generate_visualizations.py @@ -27,8 +27,8 @@ CHARTS = ASSETS / "charts" OMNI_RELAY = { "status": "pending_huggingface_gated_access", "dataset": "ropedia-ai/xperience-10m", - "relay_server": "ANGEL-A100-80Gx4", - "training_server": "ANGEL-H20-96GX8", + "staging": "prepared_generic_host_to_host_transfer", + "training_target": "external_multi_gpu_training_host", "selection_strategy": "stratified_round_robin_by_top_level_session", "target_episodes": 32, "selected_sessions": 32, @@ -37,7 +37,7 @@ OMNI_RELAY = { "estimated_bytes": 72031620552, "exclude": ["visualization.rrd"], "blocker": "Hugging Face returns 403 pending review for the full Xperience-10M gated dataset.", - "claim_boundary": "No real 32-episode fine-tune is claimed until the watcher downloads data, transfers it to H20, and the held-out evaluation runs.", + "claim_boundary": "No real 32-episode fine-tune is claimed until gated data is available locally and held-out evaluation runs.", } diff --git a/scripts/omni/build_episode_manifest.py b/scripts/omni/build_episode_manifest.py index ec529a3400cf76a8173ba11954328d233422ed7c..aa0e40a52e46662b03f1bf987dd0299aba854338 100644 --- a/scripts/omni/build_episode_manifest.py +++ b/scripts/omni/build_episode_manifest.py @@ -2,7 +2,7 @@ """Build a lightweight manifest for local Ropedia/Xperience episode folders. The manifest is intentionally metadata-only. It lets us decide how many -episodes fit on the H20 server before downloading or copying large media. +episodes fit on target storage before downloading or copying large media. """ from __future__ import annotations diff --git a/scripts/omni/discover_xperience10m_sources.py b/scripts/omni/discover_xperience10m_sources.py index 8731078571f79881f20cf1805a06e5d40089c7e5..38b5ea0af270d6e9e80472606704dbc5eb94f585 100644 --- a/scripts/omni/discover_xperience10m_sources.py +++ b/scripts/omni/discover_xperience10m_sources.py @@ -59,7 +59,7 @@ def parse_args() -> argparse.Namespace: workspace_default = Path(__file__).resolve().parents[2] parser = argparse.ArgumentParser(description="Discover Xperience-10M episode availability.") parser.add_argument("--workspace", type=Path, default=workspace_default) - parser.add_argument("--data-root", type=Path, default=Path("/home/cy/Ropedia/modelscope_data")) + parser.add_argument("--data-root", type=Path, default=Path("modelscope_data")) parser.add_argument("--output", type=Path, default=Path("results/omni_finetune/source_discovery.json")) parser.add_argument("--report-output", type=Path, default=Path("results/omni_finetune/DATA_BLOCKER_REPORT.md")) parser.add_argument("--target-episodes", type=int, default=32) @@ -257,7 +257,7 @@ def pick_source(local: dict, modelscope: dict, huggingface: dict, target: int) - blockers = [ f"Not enough degraded-valid episodes for a 32-episode pilot. Need {target}, local has {local['num_degraded_valid_episodes']}.", - "Current H20 path remains one-episode proof-of-stack only.", + "Current local path remains one-episode proof-of-stack only.", ] if local["num_episodes"] == 0: blockers.append(f"No local annotation.hdf5 found under {local.get('data_root', 'configured data root')}") diff --git a/scripts/omni/omni_finetune_runbook.py b/scripts/omni/omni_finetune_runbook.py index 0fa562d1bbc4e385ad7ce10fefb6682a26525e0f..bbaf5ccba7b6e3ff94695725ce4cf200577df694 100644 --- a/scripts/omni/omni_finetune_runbook.py +++ b/scripts/omni/omni_finetune_runbook.py @@ -48,7 +48,7 @@ def preflight_snapshot() -> dict: "nvidia_smi": command_output(["nvidia-smi", "--query-gpu=name,memory.total", "--format=csv,noheader"]), "cuda_visible_devices": command_output(["bash", "-lc", "printf %s \"${CUDA_VISIBLE_DEVICES:-unset}\""]), "ffmpeg": command_output(["ffmpeg", "-version"]).splitlines()[0], - "disk_home": command_output(["df", "-h", "/home/cy"]), + "disk_workdir": command_output(["df", "-h", "."]), } @@ -58,7 +58,7 @@ def stage_commands(run_id: str, manifest_path: Path) -> list[dict]: return [ { "phase": "phase_0_preflight", - "goal": "Confirm H20 runtime, local Qwen weights, ModelScope access, ffmpeg, and HOMIE loader.", + "goal": "Confirm accelerator runtime, local Qwen weights, dataset access, ffmpeg, and HOMIE loader.", "commands": [ "nvidia-smi", "ffmpeg -version", @@ -69,7 +69,7 @@ def stage_commands(run_id: str, manifest_path: Path) -> list[dict]: "phase": "phase_1_one_episode_smoke", "goal": "Reproduce adapter smoke and validate JSONL/media generation.", "commands": [ - f"python scripts/omni/build_episode_manifest.py --data-root /home/cy/Ropedia/modelscope_data --max-episodes 1 --output {manifest_path}", + f"python scripts/omni/build_episode_manifest.py --data-root /path/to/xperience10m_data --max-episodes 1 --output {manifest_path}", f"python scripts/omni/export_qwen3_omni_action_dataset.py --manifest {manifest_path} --max-windows-per-episode 16 --run-id {run_id}_dataset", f"python scripts/omni/qwen3_omni_inference_smoke.py --dataset-jsonl {dataset_jsonl} --sample-limit 3 --run-id {run_id}_zero_shot", ], @@ -78,7 +78,7 @@ def stage_commands(run_id: str, manifest_path: Path) -> list[dict]: "phase": "phase_2_three_episode_overfit", "goal": "Train adapter-only and Qwen LoRA on 3 episodes; require decreasing loss and >=98% JSON validity.", "commands": [ - f"python scripts/omni/build_episode_manifest.py --data-root /home/cy/Ropedia/modelscope_data --max-episodes 3 --output {manifest_path}", + f"python scripts/omni/build_episode_manifest.py --data-root /path/to/xperience10m_data --max-episodes 3 --output {manifest_path}", f"python scripts/omni/export_qwen3_omni_action_dataset.py --manifest {manifest_path} --run-id {run_id}_3ep_dataset", f"python scripts/omni/train_qwen3_omni_lora.py --dataset-jsonl results/omni_finetune/{run_id}_3ep_dataset/dataset.jsonl --run-id {run_id}_3ep_lora --max-train-samples 256", ], @@ -87,7 +87,7 @@ def stage_commands(run_id: str, manifest_path: Path) -> list[dict]: "phase": "phase_3_32_episode_pilot", "goal": "Run adapter-only, frozen Qwen, Qwen LoRA video/audio/text, and Qwen LoRA plus sensor bridge.", "commands": [ - f"python scripts/omni/build_episode_manifest.py --data-root /home/cy/Ropedia/modelscope_data --max-episodes 32 --output {manifest_path}", + f"python scripts/omni/build_episode_manifest.py --data-root /path/to/xperience10m_data --max-episodes 32 --output {manifest_path}", f"python scripts/omni/export_qwen3_omni_action_dataset.py --manifest {manifest_path} --run-id {run_id}_dataset", f"python scripts/omni/train_qwen3_omni_lora.py --dataset-jsonl {dataset_jsonl} --run-id {run_id}_lora", f"python scripts/omni/eval_qwen3_omni_lora.py --dataset-jsonl {dataset_jsonl} --adapter-dir checkpoints/{run_id}_lora/adapter_lora --run-id {run_id}_eval", diff --git a/scripts/omni/plan_finetune_sample_budget.py b/scripts/omni/plan_finetune_sample_budget.py index abd73f3cc2c58a076dfd3e14d154640e14121da7..2ca61a36614e20497a47967e50ba0937ca80290e 100644 --- a/scripts/omni/plan_finetune_sample_budget.py +++ b/scripts/omni/plan_finetune_sample_budget.py @@ -1,5 +1,5 @@ #!/usr/bin/env python3 -"""Plan Xperience-10M episode counts for an H20 fine-tuning run. +"""Plan Xperience-10M episode counts for a fine-tuning run. This is a storage and evaluation-design helper. It does not train a model and does not invent results. Use it before downloading many episodes. @@ -19,7 +19,7 @@ GB = 1024 ** 3 def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description="Estimate feasible Xperience-10M fine-tuning sample counts.") - parser.add_argument("--storage-root", type=Path, default=Path("/home/cy"), help="Disk root to inspect.") + parser.add_argument("--storage-root", type=Path, default=Path("."), help="Disk root to inspect.") parser.add_argument("--free-gb", type=float, default=None, help="Override measured free space in GiB.") parser.add_argument("--target-free-after-download-gb", type=float, default=800.0) parser.add_argument("--model-cache-gb", type=float, default=250.0) diff --git a/scripts/omni/qwen3_omni_next_steps.md b/scripts/omni/qwen3_omni_next_steps.md index 97a36d7ace83b75a17da23b59f10733c0812cb59..67261fee39aff72782875e3b3090cd73dd8baf3f 100644 --- a/scripts/omni/qwen3_omni_next_steps.md +++ b/scripts/omni/qwen3_omni_next_steps.md @@ -26,7 +26,7 @@ comparison. Assistant outputs are strict JSON with these fields: Suggested progression: -1. Phase 0: preflight H20 GPUs, CUDA/PyTorch, free disk, ModelScope, ffmpeg, +1. Phase 0: preflight accelerator runtime, CUDA/PyTorch, free disk, dataset access, ffmpeg, HOMIE, and local Qwen3-Omni-Instruct weights. 2. Phase 1: one-episode smoke with adapter-only plus JSONL/media validation. 3. Phase 2: three-episode overfit for adapter-only and Qwen LoRA. @@ -38,7 +38,7 @@ Concrete command sequence: ```bash python scripts/omni/build_episode_manifest.py \ - --data-root /home/cy/Ropedia/modelscope_data \ + --data-root /path/to/xperience10m_data \ --max-episodes 32 \ --output results/omni_finetune/episode_manifest.json diff --git a/scripts/omni/run_omni_finetune_8gpu.sh b/scripts/omni/run_omni_finetune_8gpu.sh index 622521e194bdac736c7f5bac1d9d623c500ae919..00fde82f6e385fc8dc4491c5eeb1c6562de00239 100644 --- a/scripts/omni/run_omni_finetune_8gpu.sh +++ b/scripts/omni/run_omni_finetune_8gpu.sh @@ -1,8 +1,8 @@ #!/usr/bin/env bash set -euo pipefail -WORKSPACE="${WORKSPACE:-/home/cy/Ropedia/ropedia-xperience-10m-task-suite}" -PROJECT_ROOT="${PROJECT_ROOT:-/home/cy/Ropedia}" +WORKSPACE="${WORKSPACE:-$(pwd)}" +PROJECT_ROOT="${PROJECT_ROOT:-$(dirname "${WORKSPACE}")}" VENV_PY="${VENV_PY:-$WORKSPACE/.venv/bin/python}" RUN_ID="${RUN_ID:-xperience10m_qwen3_omni_32ep}" DATA_ROOT="${DATA_ROOT:-$PROJECT_ROOT/modelscope_data}" diff --git a/scripts/omni/stage_xperience10m_from_hf.py b/scripts/omni/stage_xperience10m_from_hf.py index c5364e1a8f85f580401b6ad719af68c1cb9278eb..b2f14b190ca1f07bbc962cb0f6181b0d28103683 100644 --- a/scripts/omni/stage_xperience10m_from_hf.py +++ b/scripts/omni/stage_xperience10m_from_hf.py @@ -6,7 +6,7 @@ This downloads leaf episode folders such as: /ep1/{annotation.hdf5,fisheye_cam0.mp4,...} It intentionally excludes visualization.rrd and writes a manifest that can be -used before transferring data to the H20 training server. +used before transferring data to a training host. """ from __future__ import annotations diff --git a/scripts/omni/transfer_xperience10m_between_hosts.sh b/scripts/omni/transfer_xperience10m_between_hosts.sh new file mode 100644 index 0000000000000000000000000000000000000000..fd71554bc47505fc784efdc5ff6b05e32692bf8e --- /dev/null +++ b/scripts/omni/transfer_xperience10m_between_hosts.sh @@ -0,0 +1,17 @@ +#!/usr/bin/env bash +set -euo pipefail + +STAGING_DIR="${STAGING_DIR:?Set STAGING_DIR to the prepared Xperience-10M staging directory.}" +TRAINING_HOST="${TRAINING_HOST:?Set TRAINING_HOST to the remote training host, for example user@hostname.}" +TRAINING_DATA_ROOT="${TRAINING_DATA_ROOT:?Set TRAINING_DATA_ROOT to the remote Xperience-10M data directory.}" +REMOTE_WORKSPACE="${REMOTE_WORKSPACE:?Set REMOTE_WORKSPACE to the remote repo checkout.}" +SSH_KEY="${SSH_KEY:-$HOME/.ssh/xperience10m_transfer}" + +rsync -avP --partial --append-verify \ + --exclude "visualization.rrd" \ + -e "ssh -i ${SSH_KEY} -o BatchMode=yes -o StrictHostKeyChecking=accept-new" \ + "${STAGING_DIR}" \ + "${TRAINING_HOST}:${TRAINING_DATA_ROOT}" + +ssh -i "${SSH_KEY}" -o BatchMode=yes -o StrictHostKeyChecking=accept-new "${TRAINING_HOST}" \ + "cd ${REMOTE_WORKSPACE} && python3 scripts/omni/discover_xperience10m_sources.py --workspace ${REMOTE_WORKSPACE} --data-root ${TRAINING_DATA_ROOT} --output results/omni_finetune/source_discovery.json --report-output results/omni_finetune/DATA_BLOCKER_REPORT.md" diff --git a/scripts/omni/watch_hf_access_and_stage_xperience10m.py b/scripts/omni/watch_hf_access_and_stage_xperience10m.py index 74262a23b521370dda62c9731ee063f3cdb831c8..122e757fa79d09007c2624ed4ded598cce048111 100644 --- a/scripts/omni/watch_hf_access_and_stage_xperience10m.py +++ b/scripts/omni/watch_hf_access_and_stage_xperience10m.py @@ -1,10 +1,10 @@ #!/usr/bin/env python3 """Poll Hugging Face gated access, then stage and optionally transfer Xperience-10M. -This is intended for an A100 relay server that can reach Hugging Face while H20 -cannot. It does a cheap HEAD request against one gated file. When access is -approved, it starts the selective 32-episode staging script and then can launch -the A100->H20 transfer script. +This is intended for a staging host that can reach Hugging Face. It does a +cheap HEAD request against one gated file. When access is approved, it starts +the selective 32-episode staging script and can optionally launch a generic +host-to-host transfer script. """ from __future__ import annotations @@ -28,10 +28,10 @@ def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--repo-id", default="ropedia-ai/xperience-10m") parser.add_argument("--probe-file", default=DEFAULT_PROBE_FILE) - parser.add_argument("--local-dir", type=Path, default=Path("/mnt/kgc/chaoyue/xperience10m_hf_staging")) - parser.add_argument("--stage-script", type=Path, default=Path("/mnt/kgc/chaoyue/xperience10m_tools/stage_xperience10m_from_hf.py")) - parser.add_argument("--transfer-script", type=Path, default=Path("/mnt/kgc/chaoyue/xperience10m_tools/transfer_xperience10m_a100_to_h20.sh")) - parser.add_argument("--log-dir", type=Path, default=Path("/mnt/kgc/chaoyue/xperience10m_logs")) + parser.add_argument("--local-dir", type=Path, default=Path(os.environ.get("XPERIENCE10M_STAGE_DIR", "xperience10m_hf_staging"))) + parser.add_argument("--stage-script", type=Path, default=Path(os.environ.get("XPERIENCE10M_STAGE_SCRIPT", "scripts/omni/stage_xperience10m_from_hf.py"))) + parser.add_argument("--transfer-script", type=Path, default=Path(os.environ.get("XPERIENCE10M_TRANSFER_SCRIPT", "scripts/omni/transfer_xperience10m_between_hosts.sh"))) + parser.add_argument("--log-dir", type=Path, default=Path(os.environ.get("XPERIENCE10M_LOG_DIR", "xperience10m_logs"))) parser.add_argument("--target-episodes", type=int, default=32) parser.add_argument("--max-top-level", type=int, default=64) parser.add_argument("--workers", type=int, default=8) @@ -108,8 +108,8 @@ def main() -> int: env = os.environ.copy() if token: env["HF_TOKEN"] = token - env.setdefault("HF_HOME", "/mnt/kgc/chaoyue/hf_home") - env.setdefault("HF_HUB_CACHE", "/mnt/kgc/chaoyue/hf_cache") + env.setdefault("HF_HOME", str(args.log_dir.parent / "hf_home")) + env.setdefault("HF_HUB_CACHE", str(args.log_dir.parent / "hf_cache")) attempt = 0 while True: diff --git a/scripts/validate_publication_package.py b/scripts/validate_publication_package.py index 2a85c58402609bec85f65a3e441b54b16c8490a9..a761d1db5a79c778f43815f41445c4f7466be5c4 100644 --- a/scripts/validate_publication_package.py +++ b/scripts/validate_publication_package.py @@ -295,7 +295,7 @@ def required_assets(root: Path) -> dict[str, bool]: "results/episode_task_suite/feature_manifest.json", "results/episode_task_suite/neural_mlp/timeline_action/metrics.json", "results/omni_finetune/DATA_BLOCKER_REPORT.md", - "results/omni_finetune/A100_HF_RELAY_STATUS.md", + "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", "scripts/episode_task_suite.py", "scripts/neural_task_models.py", "scripts/build_artifact_index.py", diff --git a/scripts/validate_scope_claims.py b/scripts/validate_scope_claims.py index ad68a5a4a18661fd8d9a4f0a1887b19898092df5..37c1ce7c5cf7f09b5f4f8886d27dd870f337ee63 100644 --- a/scripts/validate_scope_claims.py +++ b/scripts/validate_scope_claims.py @@ -1,9 +1,9 @@ #!/usr/bin/env python3 """Validate that public scope claims match the actual Xperience-10M artifacts. -This guard exists because several smoke-run files retain historical `32ep` -run identifiers in their paths. Those identifiers are useful provenance, but -they must never be presented as evidence of a real 32-episode fine-tune. +This guard exists because several readiness/provenance files retain historical +`32ep` run identifiers in their paths. Those identifiers are useful provenance, +but they must never be presented as evidence of a real 32-episode fine-tune. """ from __future__ import annotations @@ -144,7 +144,7 @@ def scan_historical_result_identifiers() -> list[dict]: continue records.append( { - "classification": "historical_identifier_in_smoke_artifact", + "classification": "historical_identifier_in_readiness_artifact", "path": relative_path, "line": line_number, "patterns": matched, @@ -188,8 +188,8 @@ def build_report() -> dict: checks.append( check( "reviewer_packet_forbids_32_episode_inference", - "32-episode" in do_not_infer and "smoke run" in do_not_infer, - "reviewer packet explicitly warns not to treat the smoke run as a 32-episode fine-tune", + "32-episode" in do_not_infer and "readiness run" in do_not_infer, + "reviewer packet explicitly warns not to treat the readiness run as a 32-episode fine-tune", ["docs/data/reviewer_packet.json"], ) ) @@ -207,7 +207,7 @@ def build_report() -> dict: split_counts = dataset_manifest.get("split_counts", {}) checks.append( check( - "omni_dataset_manifest_is_smoke_only", + "omni_dataset_manifest_is_readiness_only", dataset_manifest.get("num_episodes") == 1 and dataset_manifest.get("num_samples") == 128 and split_counts == {"train": 128}, @@ -221,7 +221,7 @@ def build_report() -> dict: checks.append( check( - "omni_training_metadata_is_smoke_only", + "omni_training_metadata_is_readiness_only", training_metadata.get("num_train_samples") == 128 and training_metadata.get("num_val_samples") == 0, ( @@ -260,7 +260,7 @@ def build_report() -> dict: historical_identifiers = scan_historical_result_identifiers() checks.append( check( - "historical_32ep_identifiers_are_confined_to_smoke_artifacts", + "historical_32ep_identifiers_are_confined_to_readiness_artifacts", bool(historical_identifiers), f"historical identifiers found in result provenance files={len(historical_identifiers)}", ["results/omni_finetune/"],