Image-Text-to-Text
MLX
Safetensors
gemma4
apple-silicon
quantized
mixed-precision
axquant
axq
development
gemma-4
4bit
4-bit precision
mtp
vision
conversational
Instructions to use AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP with MLX:
# Make sure mlx-vlm is installed # pip install --upgrade mlx-vlm from mlx_vlm import load, generate from mlx_vlm.prompt_utils import apply_chat_template from mlx_vlm.utils import load_config # Load the model model, processor = load("AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP") config = load_config("AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP") # Prepare input image = ["http://images.cocodataset.org/val2017/000000039769.jpg"] prompt = "Describe this image." # Apply chat template formatted_prompt = apply_chat_template( processor, config, prompt, num_images=1 ) # Generate output output = generate(model, processor, formatted_prompt, image) print(output) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Pi
How to use AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP"
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent
How to use AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP with Hermes Agent:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP"
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 AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP"
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 "AutomatosX/AX-gemma-4-26b-a4b-MLX-AXQ-4bit-MTP" \ --custom-provider-id mlx-lm \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
| { | |
| "created_at": "2026-08-30T14:15:10.446736Z", | |
| "plan_sha256": "b7f9acf2a737308ebaa8a5182d7f27a626efa27d8590c1e887e5414255925a73", | |
| "records": [ | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.embed_tokens", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.0.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.1.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.10.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.12.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.13.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.14.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.15.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.16.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.18.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.19.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.2.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.20.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.21.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.22.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.24.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.25.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.26.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.27.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.28.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.3.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.4.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.6.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.7.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.8.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.experts.gate_up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.self_attn.k_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.self_attn.o_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.self_attn.q_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.9.self_attn.v_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.11.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.17.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.23.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 64, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.experts.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.29.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.mlp.down_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.mlp.gate_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 4, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.mlp.up_proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| }, | |
| { | |
| "bits": 8, | |
| "fallback": false, | |
| "group_size": 32, | |
| "metadata": {}, | |
| "method": "affine", | |
| "module_path": "model.language_model.layers.5.router.proj", | |
| "note": "mlx-lm affine packing", | |
| "success": true | |
| } | |
| ], | |
| "schema_version": "axquant.quantizer-execution.v1" | |
| } | |