Correct the padding-mask guidance: logit_bias, not allowed_token_ids

Correct the padding-mask guidance: logit_bias, not allowed_token_ids

Ship the modeling files on main: vLLM reads them from main at any revision

Ship the modeling files on main: vLLM reads them from main at any revision

Point the card at the hf-transformers and vllm branches

Upload folder using huggingface_hub

Upload folder using huggingface_hub

initial commit