Efficient-Large-Model / VILA

VILA - a multi-image visual language model with training, inference and evaluation recipe, deployable from cloud to edge (Jetson Orin and laptops)
Apache License 2.0
878 stars 55 forks source link

Bug in conversation.py #71

Open zhang-jr opened 1 month ago

zhang-jr commented 1 month ago

A minor bug was found here. It seems the assistantrole in conversation.py is faulty. It should be <|start_header_id|>assistant<|end_header_id|>\n\n.