Error

[CHOUDDIDIBUYAOSONG]

I want to know how to train after extracting features, the jsonl file generated by the features is different from the training data format

The jsonl file I output doesn't have prompt_len, begin_idx_list, ll_tokens_list

Do I use output.jsonl, output_1.jsonl, or both?

Can the author answer that? Thank you very much

[joyjiaxi]

i am also confused about how to proceed from gen_features to train.py, some instructions for steps in between seems like missing