423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2026-04-18 02:46:07 +08:00
Code Issues Packages Projects Releases Wiki Activity

[fix] fit neat_packing & mrope model packing (#10283)

Browse Source 
Co-authored-by: Yaowei Zheng <hiyouga@buaa.edu.cn>
This commit is contained in:
Kingsley
2026-03-20 16:50:11 +08:00
committed by GitHub
parent d91d8af89e
commit 833f6027b1
15 changed files with 520 additions and 93 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llamafactory/train/sft/trainer.py
View File

@@ -214,7 +214,7 @@ class CustomSeq2SeqTrainer(Seq2SeqTrainer):
pad_len = np.nonzero(preds[i] != self.processing_class.pad_token_id)[0]
if len(pad_len): # move pad token to last
preds[i] = np.concatenate((preds[i][pad_len[0] :], preds[i][: pad_len[0]]), axis=-1)
input_ids_column = dataset["input_ids"]
try:
input_ids_list = input_ids_column.to_pylist()