MilaNLProc / simple-generation

A python package to run inference with HuggingFace language and vision-language checkpoints wrapping many convenient features.
Other
25 stars 2 forks source link

Feat/use vlm without images #7

Closed g8a9 closed 1 month ago

g8a9 commented 1 month ago

This PR enables the possibility of using VLMs as plain language models by making image inputs optional. There are no API changes; the only change is that image inputs have become optional.