microsoft

/

Phi-3-vision-128k-instruct

like 945

Follow

Microsoft 7.72k

Extend eos_token_id with pad token because that's the token stored in processor.tokenizer.eos_token_id. That also simplifies the sample from the model card from

generate_ids = model.generate(**inputs, eos_token_id=processor.tokenizer.eos_token_id, **generation_args)

to

generate_ids = model.generate(**inputs, **generation_args)