model card: tasks / rig / provenance / training stats (auto)

README.md

@@ -1,164 +1,64 @@
 ---
-datasets: VibeCuisine/vibepi3-grab-poseexpert-curated-081826
+datasets:
+- VibeCuisine/vibepi3-grab-poseexpert-curated-081826
 library_name: lerobot
-license: apache-2.0
-model_name: act
 pipeline_tag: robotics
 tags:
-- lerobot
 - act
-- robotics
+- vibe-data-studio
+- hardware:vibeboard_v3
 ---
 
-# Model Card for act
+<!-- vibedata:model-card -->
+# vibepi3-grab-act-aug-081826
 
-<!-- Provide a quick summary of what the model is/does. -->
+act policy trained via Vibe Data Studio on 39 teleop episodes.
 
+## At a glance
 
-[Action Chunking with Transformers (ACT)](https://huggingface.co/papers/2304.13705) is an imitation-learning method that predicts short action chunks instead of single steps. It learns from teleoperated data and often achieves high success rates.
+| | |
+|---|---|
+| **Architecture** | act (from scratch) |
+| **Dataset** | [vibepi3-grab-poseexpert-curated-081826](https://huggingface.co/datasets/VibeCuisine/vibepi3-grab-poseexpert-curated-081826) @ `aa2bd373` |
+| **Scale** | 39 episodes · 2,112 frames @ 20 fps |
+| **Steps / batch / seed** | 30,000 / 8 / 1000 |
+| **lerobot** | `b6d46c40` |
+| **Provider / GPU** | site-gpu / site GPU host |
+| **Duration** | 1h 16m |
+| **Final train loss** | 0.043 |
+| **W&B** | [vds47-9284ef15](https://wandb.ai/jeremyhx-freelance/lerobot/runs/vds47-9284ef15) |
 
+## Tasks
 
+Language-conditioned — pass one of these instructions at inference:
 
+| Instruction | Episodes |
+|---|---:|
+| Grab the cucumber at the one-third point | 39 |
 
-<!-- A short demo is worth more than any description! Record a GIF/video of the policy
-running on your robot, upload it to this repo, and embed it here:
-<p align="center">
-  <img src="https://huggingface.co/<hf_user>/<policy_repo_id>/resolve/main/demo.gif" width="60%"/>
-</p>
--->
+**Episodes per task per recording rig** (derived from the nested collection manifests down to leaf recorded datasets):
 
-This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).
+| Task | vibepi3 | **Total** |
+|---|---:|---:|
+| Grab the cucumber at the one-third point | 39 | **39** |
+| **Total** | **39** | **39** |
 
-Learn how to train and run it in the [LeRobot act guide](https://huggingface.co/docs/lerobot/main/en/act), or browse the [full documentation](https://huggingface.co/docs/lerobot/index).
+## Rig
 
+- **Action/state** (7-dim): shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper, tilt
+- **Cameras** @ 20 fps: corner 640×480, top 640×480, wrist 640×480
+- **Consumed image inputs** (the checkpoint's resolved VISUAL features, from its `config.json`): `observation.images.corner`, `observation.images.top`, `observation.images.wrist`
 
----
-
-## Model Details
-
-- **License:** apache-2.0
-- **Robot type:** `vibeboard_follower_tilt`
-- **Cameras:** `corner`, `top`, `wrist`
-
-
-## Inputs & Outputs
-
-The policy consumes these observation features and produces these action features.
-
-**Inputs**
-
-| Feature | Type | Shape |
-| --- | --- | --- |
-| `observation.state` | STATE | `(7,)` |
-| `observation.images.corner` | VISUAL | `(3, 480, 640)` |
-| `observation.images.top` | VISUAL | `(3, 480, 640)` |
-| `observation.images.wrist` | VISUAL | `(3, 480, 640)` |
-
-**Outputs**
-
-| Feature | Type | Shape |
-| --- | --- | --- |
-| `action` | ACTION | `(7,)` |
-
-
-## Training Dataset
-
-- **Repository:** [VibeCuisine/vibepi3-grab-poseexpert-curated-081826](https://huggingface.co/datasets/VibeCuisine/vibepi3-grab-poseexpert-curated-081826)
-- **Episodes:** 39
-- **Frames:** 2112
-- **Frame rate:** 20 FPS
-- **Task(s):** "Grab the cucumber at the one-third point"
-
-<a class="flex" href="https://huggingface.co/spaces/lerobot/visualize_dataset?path=VibeCuisine/vibepi3-grab-poseexpert-curated-081826">
-<img class="block dark:hidden" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl.svg"/>
-<img class="hidden dark:block" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl-dark.svg"/>
-</a>
-
-
-## Training Configuration
-
-| Setting | Value |
-| --- | --- |
-| Training steps | 30000 |
-| Batch size | 8 |
-| Optimizer | adamw |
-| Learning rate | 1e-05 |
-| Seed | 1000 |
-| LeRobot version | 0.6.0 |
-
----
+<details>
+<summary><b>Reproduce</b> — exact training command</summary>
 
-## How to Get Started with the Model
-
-New to LeRobot? These guides cover the full workflow:
-
-- **[Install LeRobot](https://huggingface.co/docs/lerobot/main/en/installation)** — set up the `lerobot` package.
-- **[Hardware setup](https://huggingface.co/docs/lerobot/main/en/hardware_guide)** — assemble, wire, and calibrate your robot and cameras.
-- **[Record data & train a policy](https://huggingface.co/docs/lerobot/en/il_robots)** — the end-to-end imitation-learning walkthrough.
-- **[CLI cheat-sheet](https://huggingface.co/docs/lerobot/main/en/cheat-sheet)** — quick reference for the `lerobot-*` commands.
-
-The short version to run and train this policy:
-
-### Run the policy on your robot
-
-```bash
-lerobot-rollout \
-  --strategy.type=base \
-  --robot.type=vibeboard_follower_tilt \
-  --robot.port=<your_robot_port> \
-  --robot.cameras="{ <camera_1>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}, <camera_2>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}}" \
-  --policy.path=VibeCuisine/vibepi3-grab-act-aug-081826 \
-  --task="Grab the cucumber at the one-third point" \
-  --duration=60
 ```
-
-Replace the remaining `<...>` placeholders with your own values: `--robot.port` and the camera names/indices are specific to your machine, and the camera names must match the observation keys this policy was trained on.
-
-When `--strategy.type=base` is used the script doesn't record the episodes. Skipping duration will make the policy run indefinitely. For more information look at [rollout documentation](https://huggingface.co/docs/lerobot/main/en/inference).
-
-### Train your own policy
-
-```bash
-lerobot-train \
-  --dataset.repo_id=${HF_USER}/<dataset> \
-  --policy.type=act \
-  --output_dir=outputs/train/<policy_repo_id> \
-  --job_name=lerobot_training \
-  --policy.device=cuda \
-  --policy.repo_id=${HF_USER}/<policy_repo_id> \
-  --wandb.enable=true
+lerobot-train --policy.type=act --policy.repo_id=VibeCuisine/vibepi3-grab-act-aug-081826 --dataset.repo_id=VibeCuisine/vibepi3-grab-poseexpert-curated-081826 --dataset.revision=aa2bd3738231994b7a3ccba2918f1eacbc3a0dd2 --batch_size=8 --steps=30000 --save_freq=5000 --output_dir=outputs/train/vibepi3-grab-act-aug-081826 --job_name=vibepi3-grab-act-aug-081826 --policy.device=cuda --wandb.enable=true --wandb.disable_artifact=true --wandb.project=lerobot --wandb.run_id=vds47-9284ef15 --wandb.entity=jeremyhx-freelance --seed=1000 --dataset.eval_split=0.1 --eval_steps=600 --dataset.image_transforms.enable=true --policy.chunk_size=40 --policy.n_action_steps=40
 ```
 
-_Writes checkpoints to `outputs/train/<policy_repo_id>/checkpoints/`._
+`--num_workers` is appended host-side (clamped to the host's allocation).
 
----
-
-## Evaluation
-
-<!-- Report real-robot results here: run the policy several times per task and count the
-successes. Delete the "No evaluation results" line and fill in this table instead:
-
-| Task | Trials | Successes | Success rate |
-| ---- | ------ | --------- | ------------ |
-| pick the lego brick | 10 | 8 | 80% |
-
-Also worth noting: anything that affects difficulty (new object positions, lighting,
-distractors, a different robot of the same type, ...).
--->
-
-_No evaluation results have been provided for this policy yet._
+</details>
 
 ---
-
-## Citation
-
-If you use this policy, please cite the method linked in the description above, along with LeRobot:
-
-```bibtex
-@misc{cadene2024lerobot,
-    author = {Cadene, Remi and Alibert, Simon and Soare, Alexander and Gallouedec, Quentin and Zouitine, Adil and Palma, Steven and Kooijmans, Pepijn and Aractingi, Michel and Shukor, Mustafa and Aubakirova, Dana and Russi, Martino and Capuano, Francesco and Pascal, Caroline and Choghari, Jade and Moss, Jess and Wolf, Thomas},
-    title = {LeRobot: State-of-the-art Machine Learning for Real-World Robotics in Pytorch},
-    howpublished = "\url{https://github.com/huggingface/lerobot}",
-    year = {2024}
-}
-```
+*Card generated by Vibe Data Studio on run completion.*