Instructions to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0 # Run inference directly in the terminal: llama cli -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0 # Run inference directly in the terminal: llama cli -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0 # Run inference directly in the terminal: ./llama-cli -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0 # Run inference directly in the terminal: ./build/bin/llama-cli -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Use Docker
docker model run hf.co/Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
- LM Studio
- Jan
- vLLM
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Baekpica/MiMo-V2.6-Flash-RL-GGUF" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Baekpica/MiMo-V2.6-Flash-RL-GGUF", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
- Ollama
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with Ollama:
ollama run hf.co/Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
- Unsloth Desktop
- Pi
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with Docker Model Runner:
docker model run hf.co/Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
- Lemonade
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Run and chat with the model
lemonade run user.MiMo-V2.6-Flash-RL-GGUF-Q8_0
List all available models
lemonade list
- Hermes Agent
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use Baekpica/MiMo-V2.6-Flash-RL-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "Baekpica/MiMo-V2.6-Flash-RL-GGUF:Q8_0" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Publish original-representation MXFP4 BF16 calibration reference
Browse files- .gitattributes +4 -0
- MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf +3 -0
- MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf +3 -0
- MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf +3 -0
- MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf +3 -0
- README.md +59 -0
- SHA256SUMS +4 -0
- artifact-manifest.json +27 -0
- chat_template.jinja +96 -0
- reference-audit.json +30 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf filter=lfs diff=lfs merge=lfs -text
|
MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:829472a0e35d4bf136d7e44ecc9deaaadf1e001dab716c53bdeeecf42599559e
|
| 3 |
+
size 44499128352
|
MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:caf0e120484239878cbf389914e17f2dd9cb733dd0f9c87c35437b9e649ec551
|
| 3 |
+
size 44493179840
|
MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c92a8d8ba4eea32c8ff8ca7ba403bc4df90b3fec1b6bf5630d8672aeb2837d16
|
| 3 |
+
size 44493179840
|
MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c91e1e8489000fe5e281616b4b13c7710eaddf7ea82a6808458b26eb575c783
|
| 3 |
+
size 41422766592
|
README.md
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: XiaomiMiMo/MiMo-V2.6-Flash-RL
|
| 4 |
+
base_model_relation: quantized
|
| 5 |
+
library_name: gguf
|
| 6 |
+
pipeline_tag: text-generation
|
| 7 |
+
tags:
|
| 8 |
+
- gguf
|
| 9 |
+
- mimo_v2
|
| 10 |
+
- mxfp4
|
| 11 |
+
- calibration
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# MiMo-V2.6-Flash-RL GGUF — calibration reference
|
| 15 |
+
|
| 16 |
+
This repository holds intermediate artifacts actually generated while building [MiMo-V2.6-Flash-RL Mixed-Quant GGUF](https://huggingface.co/Baekpica/MiMo-V2.6-Flash-RL-Mixed-Quant-GGUF). The final compact mixed variant belongs in that separate repository.
|
| 17 |
+
|
| 18 |
+
The source is [XiaomiMiMo/MiMo-V2.6-Flash-RL](https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL), pinned to [`3b38d063180c3e4aed9691fdc735f3d10b266ee4`](https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL/tree/3b38d063180c3e4aed9691fdc735f3d10b266ee4).
|
| 19 |
+
|
| 20 |
+
## Reference representation
|
| 21 |
+
|
| 22 |
+
The four `MXFP4-BF16` shards preserve the original routed experts through an exact MXFP4 repack and expand source FP8 dense matrices to BF16. Control tensors are stored as F32. **This is not a full-BF16 source checkpoint or a Q8_0 baseline.** It provides original-checkpoint values for importance-matrix collection without recalibrating from the final IQ2 weights.
|
| 23 |
+
|
| 24 |
+
| Shard | Bytes |
|
| 25 |
+
|---|---:|
|
| 26 |
+
| `MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf` | 44,499,128,352 |
|
| 27 |
+
| `MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf` | 44,493,179,840 |
|
| 28 |
+
| `MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf` | 44,493,179,840 |
|
| 29 |
+
| `MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf` | 41,422,766,592 |
|
| 30 |
+
| **Total** | **174,908,254,624** |
|
| 31 |
+
|
| 32 |
+
Keep all shards together and open the first shard. The language artifact includes the checkpoint's three embedded MTP blocks; this is not evidence of validated speculative decoding. Multimodal encoders and the separate DFlash model are separate components and are not supplied by these four shards alone.
|
| 33 |
+
|
| 34 |
+
## Validation status
|
| 35 |
+
|
| 36 |
+
- All 90 downloaded source repository files passed Hub checksum verification.
|
| 37 |
+
- Independent MXFP4 repacking and tensor-parallel QKV ordering checks passed.
|
| 38 |
+
- All 36,096 expert matrices were covered by a deterministic **sampled-row** audit: 108,257 rows matched the independently repacked source. See `reference-audit.json`; this is not an exhaustive payload comparison.
|
| 39 |
+
- The reference loaded on a B300 GPU and passed a short arithmetic decode check.
|
| 40 |
+
- Original-representation text imatrix collection is running as of 2026-09-22.
|
| 41 |
+
- Final mixed-model quality, multimodal end-to-end behavior, MTP/DFlash execution, and DGX Spark serving remain pending. No throughput or benchmark qualification is claimed here.
|
| 42 |
+
|
| 43 |
+
`artifact-manifest.json` records exact file sizes and SHA-256 digests. `SHA256SUMS` can be checked after download:
|
| 44 |
+
|
| 45 |
+
```bash
|
| 46 |
+
hf download Baekpica/MiMo-V2.6-Flash-RL-GGUF --local-dir ./MiMo-reference
|
| 47 |
+
cd ./MiMo-reference
|
| 48 |
+
sha256sum -c SHA256SUMS
|
| 49 |
+
```
|
| 50 |
+
|
| 51 |
+
## Chat template and reproduction
|
| 52 |
+
|
| 53 |
+
`chat_template.jinja` is copied byte for byte from the pinned source. Use the original MiMo tokenizer and special-token mapping. Correct image, audio, and video processing additionally requires the corresponding native encoder and input protocol.
|
| 54 |
+
|
| 55 |
+
The conversion uses a local MXFP4 extension to llama.cpp revision `5836771`. Reproduction scripts will accompany the mixed release and private Spark handoff. The reference's approximately 175 GB file size is not the compact DGX Spark target; see the separate mixed model card for that recipe and its current estimates.
|
| 56 |
+
|
| 57 |
+
## License
|
| 58 |
+
|
| 59 |
+
MIT, inherited from the pinned upstream model.
|
SHA256SUMS
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
829472a0e35d4bf136d7e44ecc9deaaadf1e001dab716c53bdeeecf42599559e MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf
|
| 2 |
+
caf0e120484239878cbf389914e17f2dd9cb733dd0f9c87c35437b9e649ec551 MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf
|
| 3 |
+
c92a8d8ba4eea32c8ff8ca7ba403bc4df90b3fec1b6bf5630d8672aeb2837d16 MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf
|
| 4 |
+
5c91e1e8489000fe5e281616b4b13c7710eaddf7ea82a6808458b26eb575c783 MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf
|
artifact-manifest.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"source": "XiaomiMiMo/MiMo-V2.6-Flash-RL",
|
| 3 |
+
"source_revision": "3b38d063180c3e4aed9691fdc735f3d10b266ee4",
|
| 4 |
+
"role": "Original MXFP4 repack / FP8-expanded BF16 calibration reference",
|
| 5 |
+
"files": [
|
| 6 |
+
{
|
| 7 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf",
|
| 8 |
+
"bytes": 44499128352,
|
| 9 |
+
"sha256": "829472a0e35d4bf136d7e44ecc9deaaadf1e001dab716c53bdeeecf42599559e"
|
| 10 |
+
},
|
| 11 |
+
{
|
| 12 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf",
|
| 13 |
+
"bytes": 44493179840,
|
| 14 |
+
"sha256": "caf0e120484239878cbf389914e17f2dd9cb733dd0f9c87c35437b9e649ec551"
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf",
|
| 18 |
+
"bytes": 44493179840,
|
| 19 |
+
"sha256": "c92a8d8ba4eea32c8ff8ca7ba403bc4df90b3fec1b6bf5630d8672aeb2837d16"
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf",
|
| 23 |
+
"bytes": 41422766592,
|
| 24 |
+
"sha256": "5c91e1e8489000fe5e281616b4b13c7710eaddf7ea82a6808458b26eb575c783"
|
| 25 |
+
}
|
| 26 |
+
]
|
| 27 |
+
}
|
chat_template.jinja
ADDED
|
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- macro render_value(value) -%}
|
| 2 |
+
{%- if value is string -%}
|
| 3 |
+
{{- value -}}
|
| 4 |
+
{%- else -%}
|
| 5 |
+
{{- value | tojson(ensure_ascii=False) -}}
|
| 6 |
+
{%- endif -%}
|
| 7 |
+
{%- endmacro -%}
|
| 8 |
+
|
| 9 |
+
{%- macro render_content(message_content) -%}
|
| 10 |
+
{%- if message_content is string -%}
|
| 11 |
+
{{- message_content -}}
|
| 12 |
+
{%- elif message_content is iterable -%}
|
| 13 |
+
{%- for part in message_content -%}
|
| 14 |
+
{%- if part is not mapping -%}
|
| 15 |
+
{{- part -}}
|
| 16 |
+
{%- elif part['type'] == 'image' or 'image' in part or 'image_url' in part -%}
|
| 17 |
+
{{- '<|vision_start|><|image_pad|><|vision_end|>' -}}
|
| 18 |
+
{%- elif part['type'] == 'audio' or part['type'] == 'input_audio' or 'audio' in part or 'audio_url' in part or 'input_audio' in part -%}
|
| 19 |
+
{{- '<|mimo_audio_start|><|audio_pad|><|mimo_audio_end|>' -}}
|
| 20 |
+
{%- elif part['type'] == 'video' or 'video' in part or 'video_url' in part -%}
|
| 21 |
+
{{- '<|vision_start|><|video_pad|><|vision_end|>' -}}
|
| 22 |
+
{%- elif 'text' in part -%}
|
| 23 |
+
{{- part['text'] -}}
|
| 24 |
+
{%- endif -%}
|
| 25 |
+
{%- endfor -%}
|
| 26 |
+
{%- endif -%}
|
| 27 |
+
{%- endmacro -%}
|
| 28 |
+
|
| 29 |
+
{%- macro render_tools(tools) -%}
|
| 30 |
+
{{- 'You are provided with the following tools:\n\n<tools>' -}}
|
| 31 |
+
{%- for tool in tools -%}
|
| 32 |
+
{{- '\n' ~ (tool | tojson(ensure_ascii=False)) -}}
|
| 33 |
+
{%- endfor -%}
|
| 34 |
+
{{- '\n</tools>' -}}
|
| 35 |
+
{%- endmacro -%}
|
| 36 |
+
|
| 37 |
+
{%- macro render_tool_calls(tool_calls) -%}
|
| 38 |
+
{%- for tool_call in tool_calls -%}
|
| 39 |
+
{%- if tool_call.function is defined -%}
|
| 40 |
+
{%- set tool_call = tool_call.function -%}
|
| 41 |
+
{%- elif tool_call.custom is defined -%}
|
| 42 |
+
{%- set tool_call = tool_call.custom -%}
|
| 43 |
+
{%- endif -%}
|
| 44 |
+
{{- '<tool_call><function=' ~ tool_call.name ~ '>' -}}
|
| 45 |
+
{%- if tool_call.input is defined and tool_call.input is string -%}
|
| 46 |
+
{{- tool_call.input -}}
|
| 47 |
+
{%- elif tool_call.arguments -%}
|
| 48 |
+
{%- if tool_call.arguments is string -%}
|
| 49 |
+
{{- tool_call.arguments -}}
|
| 50 |
+
{%- else -%}
|
| 51 |
+
{%- for args_name, args_value in tool_call.arguments | items -%}
|
| 52 |
+
{{- '<parameter=' ~ args_name ~ '>' ~ render_value(args_value) ~ '</parameter>' -}}
|
| 53 |
+
{%- endfor -%}
|
| 54 |
+
{%- endif -%}
|
| 55 |
+
{%- endif -%}
|
| 56 |
+
{{- '</function></tool_call>' -}}
|
| 57 |
+
{%- endfor -%}
|
| 58 |
+
{%- endmacro -%}
|
| 59 |
+
|
| 60 |
+
{%- macro render_assistant_message(message) -%}
|
| 61 |
+
{%- set content = render_content(message.content) -%}
|
| 62 |
+
{%- set reasoning = message.reasoning_content if message.reasoning_content is string else '' -%}
|
| 63 |
+
{{- '<|im_start|>assistant\n<think>' ~ reasoning ~ '</think>' ~ content -}}
|
| 64 |
+
{%- if message.tool_calls is defined and message.tool_calls is iterable and message.tool_calls | length > 0 -%}
|
| 65 |
+
{{- render_tool_calls(message.tool_calls) -}}
|
| 66 |
+
{%- endif -%}
|
| 67 |
+
{{- '<|im_end|>' -}}
|
| 68 |
+
{%- endmacro -%}
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
{%- if tools is defined and tools is iterable and tools | length > 0 -%}
|
| 72 |
+
{{- '<|im_start|>system\n' ~ render_tools(tools) ~ '<|im_end|>' -}}
|
| 73 |
+
{%- endif -%}
|
| 74 |
+
|
| 75 |
+
{%- for message in messages -%}
|
| 76 |
+
{%- if message.role == 'assistant' -%}
|
| 77 |
+
{{- render_assistant_message(message) -}}
|
| 78 |
+
{%- else -%}
|
| 79 |
+
{%- set body = render_content(message.content) -%}
|
| 80 |
+
{{- '<|im_start|>' ~ message.role ~ '\n' ~ body -}}
|
| 81 |
+
{%- if message.tools is defined and message.tools is iterable and message.tools | length > 0 -%}
|
| 82 |
+
{%- if body -%}
|
| 83 |
+
{{- '\n\n' -}}
|
| 84 |
+
{%- endif -%}
|
| 85 |
+
{{- render_tools(message.tools) -}}
|
| 86 |
+
{%- endif -%}
|
| 87 |
+
{{- '<|im_end|>' -}}
|
| 88 |
+
{%- endif -%}
|
| 89 |
+
{%- endfor -%}
|
| 90 |
+
|
| 91 |
+
{%- if add_generation_prompt -%}
|
| 92 |
+
{{- '<|im_start|>assistant\n' -}}
|
| 93 |
+
{%- if enable_thinking is false -%}
|
| 94 |
+
{{- '<think></think>' -}}
|
| 95 |
+
{%- endif -%}
|
| 96 |
+
{%- endif -%}
|
reference-audit.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"reference_files": [
|
| 3 |
+
{
|
| 4 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00001-of-00004.gguf",
|
| 5 |
+
"bytes": 44499128352,
|
| 6 |
+
"tensors": 39
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00002-of-00004.gguf",
|
| 10 |
+
"bytes": 44493179840,
|
| 11 |
+
"tensors": 39
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00003-of-00004.gguf",
|
| 15 |
+
"bytes": 44493179840,
|
| 16 |
+
"tensors": 39
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"file": "MiMo-V2.6-Flash-RL-MXFP4-BF16-00004-of-00004.gguf",
|
| 20 |
+
"bytes": 41422766592,
|
| 21 |
+
"tensors": 391
|
| 22 |
+
}
|
| 23 |
+
],
|
| 24 |
+
"expert_matrices": 36096,
|
| 25 |
+
"sampled_rows": 108257,
|
| 26 |
+
"all_experts_checked": true,
|
| 27 |
+
"errors": [],
|
| 28 |
+
"seconds": 44.81047034263611,
|
| 29 |
+
"passed": true
|
| 30 |
+
}
|