How to use soundTeam/Q3-8B-Kintsugi_mlx-6bpw with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("soundTeam/Q3-8B-Kintsugi_mlx-6bpw") prompt = "Write a story about Einstein" messages = [{"role": "user", "content": prompt}] prompt = tokenizer.apply_chat_template( messages, add_generation_prompt=True ) text = generate(model, tokenizer, prompt=prompt, verbose=True)
How to use soundTeam/Q3-8B-Kintsugi_mlx-6bpw with Unsloth Studio:
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for soundTeam/Q3-8B-Kintsugi_mlx-6bpw to start chatting
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for soundTeam/Q3-8B-Kintsugi_mlx-6bpw to start chatting
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for soundTeam/Q3-8B-Kintsugi_mlx-6bpw to start chatting
pip install unsloth from unsloth import FastModel model, tokenizer = FastModel.from_pretrained( model_name="soundTeam/Q3-8B-Kintsugi_mlx-6bpw", max_seq_length=2048, )
How to use soundTeam/Q3-8B-Kintsugi_mlx-6bpw with MLX LM:
# Install MLX LM uv tool install mlx-lm # Interactive chat REPL mlx_lm.chat --model "soundTeam/Q3-8B-Kintsugi_mlx-6bpw"
# Install MLX LM uv tool install mlx-lm # Start the server mlx_lm.server --model "soundTeam/Q3-8B-Kintsugi_mlx-6bpw" # Calling the OpenAI-compatible server with curl curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "soundTeam/Q3-8B-Kintsugi_mlx-6bpw", "messages": [ {"role": "user", "content": "Hello"} ] }'
