Instructions to use wejoncy/gemma-4-E4B-fp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use wejoncy/gemma-4-E4B-fp8 with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("wejoncy/gemma-4-E4B-fp8") model = AutoModelForMultimodalLM.from_pretrained("wejoncy/gemma-4-E4B-fp8") - Notebooks
- Google Colab
- Kaggle
| { | |
| "producer": { | |
| "name": "modelopt", | |
| "version": "0.33.0" | |
| }, | |
| "quantization": { | |
| "quant_algo": "FP8", | |
| "kv_cache_quant_algo": null, | |
| "exclude_modules": [ | |
| "model.layers.audio_tower*", | |
| "model.layers.embed_audio*", | |
| "model.layers.embed_vision*", | |
| "model.layers.vision_tower*", | |
| "lm_head" | |
| ] | |
| } | |
| } |