Francis Chen

https://franciszchen.github.io/

Yale University New Haven Senior Postdoc at Yale University; Assistant professor at HKISI-CAS

Results 2 comments of


                                            Francis Chen

A decoder-only architecture is being used, but right-padding was detected! For correct generation results, please set `padding_side='left'` when initializing the tokenizer.

Same question about the padding_side warning. Have you solved this problem?

Could I have the stage 1 checkpoint model?

Same question. It would be beneficial to perform the stage 2 directly, without repeating the burdensome stage 1 for users.