Instructions to use mlx-community/diffusiongemma-26B-A4B-it-nvfp4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/diffusiongemma-26B-A4B-it-nvfp4 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("mlx-community/diffusiongemma-26B-A4B-it-nvfp4") config = load_config("mlx-community/diffusiongemma-26B-A4B-it-nvfp4") # 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 mlx-community/diffusiongemma-26B-A4B-it-nvfp4 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/diffusiongemma-26B-A4B-it-nvfp4"
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": "mlx-community/diffusiongemma-26B-A4B-it-nvfp4" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent
How to use mlx-community/diffusiongemma-26B-A4B-it-nvfp4 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/diffusiongemma-26B-A4B-it-nvfp4"
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/diffusiongemma-26B-A4B-it-nvfp4
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use mlx-community/diffusiongemma-26B-A4B-it-nvfp4 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/diffusiongemma-26B-A4B-it-nvfp4"
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/diffusiongemma-26B-A4B-it-nvfp4" \ --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"
Download config.json from mlx-community/diffusiongemma-26B-A4B-it-nvfp4: direct link, hf CLI and curl.
- Browser
- Download file 31.8 kB
-
https://huggingface.co/mlx-community/diffusiongemma-26B-A4B-it-nvfp4/resolve/main/config.json
- Command line
-
hf download hf://mlx-community/diffusiongemma-26B-A4B-it-nvfp4/config.json
-
curl -L -o config.json https://huggingface.co/mlx-community/diffusiongemma-26B-A4B-it-nvfp4/resolve/main/config.json
31.8 kB
{
"architectures": [
"DiffusionGemmaForBlockDiffusion"
],
"boi_token_id": 255999,
"canvas_length": 256,
"dtype": "bfloat16",
"eoi_token_id": 258882,
"eos_token_id": [
1,
106,
50
],
"generation_config": {
"confidence_threshold": 0.005,
"eos_token_id": [
1,
106,
50
],
"max_denoising_steps": 48,
"max_new_tokens": 256,
"pad_token_id": 0,
"sampler_config": {
"_cls_name": "EntropyBoundSamplerConfig",
"entropy_bound": 0.1
},
"stability_threshold": 1,
"t_max": 0.8,
"t_min": 0.4,
"transformers_version": "5.8.0.dev0"
},
"image_token_id": 258880,
"initializer_range": 0.02,
"model_type": "diffusion_gemma",
"quantization": {
"group_size": 16,
"bits": 4,
"mode": "nvfp4",
"model.decoder.layers.0.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.router.proj": {
"group_size": 64,
"bits": 8
}
},
"quantization_config": {
"group_size": 16,
"bits": 4,
"mode": "nvfp4",
"model.decoder.layers.0.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.4.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.5.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.6.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.7.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.8.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.9.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.10.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.11.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.12.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.13.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.14.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.15.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.16.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.17.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.18.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.19.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.20.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.21.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.22.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.23.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.24.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.25.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.26.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.27.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.28.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.29.router.proj": {
"group_size": 64,
"bits": 8
}
},
"text_config": {
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 2,
"dtype": "bfloat16",
"eos_token_id": 1,
"final_logit_softcapping": 30.0,
"global_head_dim": 512,
"head_dim": 256,
"hidden_activation": "gelu_pytorch_tanh",
"hidden_size": 2816,
"initializer_range": 0.02,
"intermediate_size": 2112,
"layer_types": [
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention"
],
"max_position_embeddings": 262144,
"model_type": "diffusion_gemma_text",
"moe_intermediate_size": 704,
"num_attention_heads": 16,
"num_experts": 128,
"num_global_key_value_heads": 2,
"num_hidden_layers": 30,
"num_key_value_heads": 8,
"pad_token_id": 0,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"full_attention": {
"partial_rotary_factor": 0.25,
"rope_theta": 1000000.0,
"rope_type": "proportional"
},
"sliding_attention": {
"rope_theta": 10000.0,
"rope_type": "default"
}
},
"sliding_window": 1024,
"tie_word_embeddings": true,
"top_k_experts": 8,
"use_bidirectional_attention": "vision",
"vocab_size": 262144
},
"tie_word_embeddings": true,
"transformers_version": "5.8.0.dev0",
"vision_config": {
"_name_or_path": "",
"architectures": null,
"attention_bias": false,
"attention_dropout": 0.0,
"chunk_size_feed_forward": 0,
"default_output_length": 280,
"dtype": "bfloat16",
"global_head_dim": 72,
"head_dim": 72,
"hidden_activation": "gelu_pytorch_tanh",
"hidden_size": 1152,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"initializer_range": 0.02,
"intermediate_size": 4304,
"is_encoder_decoder": false,
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"max_position_embeddings": 131072,
"model_type": "gemma4_vision",
"num_attention_heads": 16,
"num_hidden_layers": 27,
"num_key_value_heads": 16,
"output_attentions": false,
"output_hidden_states": false,
"patch_size": 16,
"pooling_kernel_size": 3,
"position_embedding_size": 10240,
"problem_type": null,
"return_dict": true,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 100.0,
"rope_type": "default"
},
"standardize": true,
"use_clipped_linears": false
},
"vision_soft_tokens_per_image": 280
}
{
"architectures": [
"DiffusionGemmaForBlockDiffusion"
],
"boi_token_id": 255999,
"canvas_length": 256,
"dtype": "bfloat16",
"eoi_token_id": 258882,
"eos_token_id": [
1,
106,
50
],
"generation_config": {
"confidence_threshold": 0.005,
"eos_token_id": [
1,
106,
50
],
"max_denoising_steps": 48,
"max_new_tokens": 256,
"pad_token_id": 0,
"sampler_config": {
"_cls_name": "EntropyBoundSamplerConfig",
"entropy_bound": 0.1
},
"stability_threshold": 1,
"t_max": 0.8,
"t_min": 0.4,
"transformers_version": "5.8.0.dev0"
},
"image_token_id": 258880,
"initializer_range": 0.02,
"model_type": "diffusion_gemma",
"quantization": {
"group_size": 16,
"bits": 4,
"mode": "nvfp4",
"model.decoder.layers.0.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.0.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.1.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.2.router.proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.gate_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.up_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.mlp.down_proj": {
"group_size": 64,
"bits": 8
},
"model.decoder.layers.3.router.proj": {