You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@PlutoXN
There are instructions in Readme.md for obtaining the checkpoints, you need first to fill a form and get the original llama model weights, then add shikra_delta_7b to the original LLaMa weights through the provided script in Readme.md
The "assert has_image" in approximately 156 lines of code is triggered in "mllm/dataset/single_image convsion. py".
The text was updated successfully, but these errors were encountered: