Text Generation
Safetensors
GGUF
English
Chinese
vllm
mimo_v2
nvfp4
modelopt
mixture-of-experts
lossless-weight-transcode
conversational
custom_code
8-bit precision
Instructions to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 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 ProCreations/MiMo-V2.6-Flash-RL-NVFP4 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 ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16 # Run inference directly in the terminal: llama cli -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16 # Run inference directly in the terminal: llama cli -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16 # Run inference directly in the terminal: ./build/bin/llama cli -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
Use Docker
docker model run hf.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
- Llama new
- LM Studio
- Jan
- vLLM
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "ProCreations/MiMo-V2.6-Flash-RL-NVFP4" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ProCreations/MiMo-V2.6-Flash-RL-NVFP4", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
- Ollama
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with Ollama:
ollama run hf.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
- Unsloth Desktop
- Pi
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
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:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with Docker Model Runner:
docker model run hf.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
- Lemonade
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
Run and chat with the model
lemonade run user.MiMo-V2.6-Flash-RL-NVFP4-BF16
List all available models
lemonade list
- Hermes Agent
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
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://127.0.0.1:9931/v1 hermes config set model.default ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use ProCreations/MiMo-V2.6-Flash-RL-NVFP4 with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16
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://127.0.0.1:9931/v1 \ --custom-model-id "ProCreations/MiMo-V2.6-Flash-RL-NVFP4:BF16" \ --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"
Download reports/heldout-nvfp4-strict.json from ProCreations/MiMo-V2.6-Flash-RL-NVFP4: direct link, hf CLI and curl.
- Browser
- Download file 1.76 kB
-
https://huggingface.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4/resolve/main/reports/heldout-nvfp4-strict.json
- Command line
-
hf download hf://ProCreations/MiMo-V2.6-Flash-RL-NVFP4/reports/heldout-nvfp4-strict.json
-
curl -L -o heldout-nvfp4-strict.json https://huggingface.co/ProCreations/MiMo-V2.6-Flash-RL-NVFP4/resolve/main/reports/heldout-nvfp4-strict.json
1.76 kB
| { | |
| "label": "nvfp4-strict", | |
| "status": "passed", | |
| "exact_runtime_gate": true, | |
| "reference_logprob_encoding": "llama.cpp uint16 with max_logit-16 clipping; KL is an approximate diagnostic. PPL ratios use the original unclipped baseline PPL, reported to 4 decimals.", | |
| "domains": [ | |
| { | |
| "domain": "code", | |
| "log": "/workspace/mimo/logs/kl-nvfp4-strict-code.log", | |
| "ppl": 2.37743, | |
| "encoded_baseline_ppl": 2.365343, | |
| "encoded_baseline_ppl_ratio": 1.00511, | |
| "mean_kl": 0.0, | |
| "max_kl": 6.3e-05, | |
| "top1_agreement_percent": 100.0, | |
| "baseline_ppl": 2.3774, | |
| "ppl_ratio": 1.0000126188272902 | |
| }, | |
| { | |
| "domain": "math", | |
| "log": "/workspace/mimo/logs/kl-nvfp4-strict-math.log", | |
| "ppl": 1.517172, | |
| "encoded_baseline_ppl": 1.515299, | |
| "encoded_baseline_ppl_ratio": 1.001236, | |
| "mean_kl": -0.0, | |
| "max_kl": 6.4e-05, | |
| "top1_agreement_percent": 100.0, | |
| "baseline_ppl": 1.5172, | |
| "ppl_ratio": 0.9999815449512258 | |
| }, | |
| { | |
| "domain": "english", | |
| "log": "/workspace/mimo/logs/kl-nvfp4-strict-english.log", | |
| "ppl": 9.05608, | |
| "encoded_baseline_ppl": 8.829277, | |
| "encoded_baseline_ppl_ratio": 1.025688, | |
| "mean_kl": -0.0, | |
| "max_kl": 6.3e-05, | |
| "top1_agreement_percent": 99.994, | |
| "baseline_ppl": 9.0561, | |
| "ppl_ratio": 0.9999977915438212 | |
| }, | |
| { | |
| "domain": "chinese", | |
| "log": "/workspace/mimo/logs/kl-nvfp4-strict-chinese.log", | |
| "ppl": 12.635223, | |
| "encoded_baseline_ppl": 12.441527, | |
| "encoded_baseline_ppl_ratio": 1.015568, | |
| "mean_kl": -0.0, | |
| "max_kl": 6.4e-05, | |
| "top1_agreement_percent": 99.988, | |
| "baseline_ppl": 12.6352, | |
| "ppl_ratio": 1.0000018203115106 | |
| } | |
| ] | |
| } | |