How to use from
Pi
Start the llama.cpp server
# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf michaelw9999/BLS-Mini-Code-1.0:NVFP4
Configure the model in Pi
# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "michaelw9999/BLS-Mini-Code-1.0:NVFP4"
        }
      ]
    }
  }
}
Run Pi
# Start Pi in your project directory:
pi
Quick Links

This is an experimental quantization of CohereLab's BLS-Mini-Code-1.0, a 30B A3B MoE model, and is still in development.
I am not affiliated with CohereLabs in any way and this is my own independent work.
To use these GGUFs with llama.cpp, you must install the llama.cpp PR #24260 for the Cohere2-MoE :

git fetch origin pull/24260/head:cohere2-moe
git checkout cohere2-moe

OR

gh pr checkout 24260
Downloads last month
436
GGUF
Model size
30B params
Architecture
cohere2moe
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for michaelw9999/BLS-Mini-Code-1.0

Quantized
(5)
this model