Text Generation
Safetensors
Model Optimizer
gemma4
nvidia
ModelOpt
Gemma-4-31B-IT
lighthouse
quantized
NVFP4
conversational
modelopt
Instructions to use nvidia/Gemma-4-31B-IT-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Inference
Download hf_quant_config.json from nvidia/Gemma-4-31B-IT-NVFP4: direct link, hf CLI and curl.
- Browser
- Download file 3.75 kB
-
https://huggingface.co/nvidia/Gemma-4-31B-IT-NVFP4/resolve/main/hf_quant_config.json
- Command line
-
hf download hf://nvidia/Gemma-4-31B-IT-NVFP4/hf_quant_config.json
-
curl -L -o hf_quant_config.json https://huggingface.co/nvidia/Gemma-4-31B-IT-NVFP4/resolve/main/hf_quant_config.json
3.75 kB
| { | |
| "producer": { | |
| "name": "modelopt", | |
| "version": "0.37.0" | |
| }, | |
| "quantization": { | |
| "quant_algo": "NVFP4", | |
| "kv_cache_quant_algo": "FP8", | |
| "group_size": 16, | |
| "exclude_modules": [ | |
| "lm_head", | |
| "model.embed_vision*", | |
| "model.language_model.layers.0.self_attn*", | |
| "model.language_model.layers.1.self_attn*", | |
| "model.language_model.layers.10.self_attn*", | |
| "model.language_model.layers.11.self_attn*", | |
| "model.language_model.layers.12.self_attn*", | |
| "model.language_model.layers.13.self_attn*", | |
| "model.language_model.layers.14.self_attn*", | |
| "model.language_model.layers.15.self_attn*", | |
| "model.language_model.layers.16.self_attn*", | |
| "model.language_model.layers.17.self_attn*", | |
| "model.language_model.layers.18.self_attn*", | |
| "model.language_model.layers.19.self_attn*", | |
| "model.language_model.layers.2.self_attn*", | |
| "model.language_model.layers.20.self_attn*", | |
| "model.language_model.layers.21.self_attn*", | |
| "model.language_model.layers.22.self_attn*", | |
| "model.language_model.layers.23.self_attn*", | |
| "model.language_model.layers.24.self_attn*", | |
| "model.language_model.layers.25.self_attn*", | |
| "model.language_model.layers.26.self_attn*", | |
| "model.language_model.layers.27.self_attn*", | |
| "model.language_model.layers.28.self_attn*", | |
| "model.language_model.layers.29.self_attn*", | |
| "model.language_model.layers.3.self_attn*", | |
| "model.language_model.layers.30.self_attn*", | |
| "model.language_model.layers.31.self_attn*", | |
| "model.language_model.layers.32.self_attn*", | |
| "model.language_model.layers.33.self_attn*", | |
| "model.language_model.layers.34.self_attn*", | |
| "model.language_model.layers.35.self_attn*", | |
| "model.language_model.layers.36.self_attn*", | |
| "model.language_model.layers.37.self_attn*", | |
| "model.language_model.layers.38.self_attn*", | |
| "model.language_model.layers.39.self_attn*", | |
| "model.language_model.layers.4.self_attn*", | |
| "model.language_model.layers.40.self_attn*", | |
| "model.language_model.layers.41.self_attn*", | |
| "model.language_model.layers.42.self_attn*", | |
| "model.language_model.layers.43.self_attn*", | |
| "model.language_model.layers.44.self_attn*", | |
| "model.language_model.layers.45.self_attn*", | |
| "model.language_model.layers.46.self_attn*", | |
| "model.language_model.layers.47.self_attn*", | |
| "model.language_model.layers.48.self_attn*", | |
| "model.language_model.layers.49.self_attn*", | |
| "model.language_model.layers.5.self_attn*", | |
| "model.language_model.layers.50.self_attn*", | |
| "model.language_model.layers.51.self_attn*", | |
| "model.language_model.layers.52.self_attn*", | |
| "model.language_model.layers.53.self_attn*", | |
| "model.language_model.layers.54.self_attn*", | |
| "model.language_model.layers.55.self_attn*", | |
| "model.language_model.layers.56.self_attn*", | |
| "model.language_model.layers.57.self_attn*", | |
| "model.language_model.layers.58.self_attn*", | |
| "model.language_model.layers.59.self_attn*", | |
| "model.language_model.layers.6.self_attn*", | |
| "model.language_model.layers.7.self_attn*", | |
| "model.language_model.layers.8.self_attn*", | |
| "model.language_model.layers.9.self_attn*", | |
| "model.vision_tower*" | |
| ] | |
| } | |
| } |