Instructions to use google/gemma-4-E2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use google/gemma-4-E2B with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("google/gemma-4-E2B") model = AutoModelForMultimodalLM.from_pretrained("google/gemma-4-E2B") - Notebooks
- Google Colab
- Kaggle
From 896x896 fixed aspect ratio to variable aspect ratio.
#3
by WillyCodesInit - opened
This right here is a game changer "Images can be processed at variable aspect ratios and resolutions", considering how gemma 3 vision encoder was handling aspect ratios.
Hi @WillyCodesInit ,
Thanks for the feedback! We are glad you found the variable aspect ratio support impactful. We appreciate you sharing your thoughts.