Text Generation
MLX
Safetensors
English
llada2_moe
optiq
diffusion
conversational
custom_code
4-bit precision
Instructions to use mlx-community/LLaDA2.2-flash-OptiQ-2bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/LLaDA2.2-flash-OptiQ-2bit 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("mlx-community/LLaDA2.2-flash-OptiQ-2bit") 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) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Pi
How to use mlx-community/LLaDA2.2-flash-OptiQ-2bit with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/LLaDA2.2-flash-OptiQ-2bit"
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "mlx-community/LLaDA2.2-flash-OptiQ-2bit" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use mlx-community/LLaDA2.2-flash-OptiQ-2bit 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 "mlx-community/LLaDA2.2-flash-OptiQ-2bit"
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 mlx-community/LLaDA2.2-flash-OptiQ-2bit
Run Hermes
hermes
- OpenClaw new
How to use mlx-community/LLaDA2.2-flash-OptiQ-2bit with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/LLaDA2.2-flash-OptiQ-2bit"
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 "mlx-community/LLaDA2.2-flash-OptiQ-2bit" \ --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"
- MLX LM
How to use mlx-community/LLaDA2.2-flash-OptiQ-2bit with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Interactive chat REPL mlx_lm.chat --model "mlx-community/LLaDA2.2-flash-OptiQ-2bit"
Run an OpenAI-compatible server
# Install MLX LM uv tool install mlx-lm # Start the server mlx_lm.server --model "mlx-community/LLaDA2.2-flash-OptiQ-2bit" # Calling the OpenAI-compatible server with curl curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "mlx-community/LLaDA2.2-flash-OptiQ-2bit", "messages": [ {"role": "user", "content": "Hello"} ] }'
| { | |
| "method": "static_mixed_precision", | |
| "base_model": "/Users/codelion/.cache/huggingface/hub/models--inclusionAI--LLaDA2.2-flash/snapshots/3adc2d7e65429b297b994ecea905f982e65517d4/", | |
| "reference": "structural_rules", | |
| "target_bpw": 2.5, | |
| "achieved_bpw": 2.508299308561932, | |
| "n_high_bits": 183, | |
| "n_low_bits": 71, | |
| "threshold": 0.0, | |
| "per_layer": { | |
| "lm_head": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.31.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.30.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.29.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.28.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.27.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.26.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.25.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.24.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.23.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.22.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.21.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.20.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.19.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.18.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.17.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.16.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.15.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.14.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.13.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.12.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.11.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.10.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.9.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.8.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.7.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.6.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.5.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.down_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.up_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.gate_proj": { | |
| "bits": 2, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.4.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.3.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.2.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.shared_experts.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.shared_experts.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.shared_experts.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.1.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.0.mlp.down_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.0.mlp.up_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.0.mlp.gate_proj": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.0.attention.dense": { | |
| "bits": 4, | |
| "group_size": 64 | |
| }, | |
| "model.layers.0.attention.query_key_value": { | |
| "bits": 4, | |
| "group_size": 64 | |
| } | |
| } | |
| } |