ros/LLaMA-Factory
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

tiny fix

Browse Source 
Former-commit-id: d8ddd07c2ed14d871fb25743c20265fc99e3e221
This commit is contained in:
hiyouga
2024-10-08 17:48:56 +08:00
parent 9df626dc18
commit b2dc6dc59a
8 changed files with 12 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llamafactory/model/model_utils/packing.py
View File

@@ -114,7 +114,7 @@ def get_unpad_data(attention_mask: "torch.Tensor") -> Tuple["torch.Tensor", "tor
def _patch_for_block_diag_attn(model_type: str) -> None:
require_version("transformers>=4.41.2,<=4.45.0", "To fix: pip install transformers>=4.41.2,<=4.45.0")
require_version("transformers>=4.41.2,<=4.45.2", "To fix: pip install transformers>=4.41.2,<=4.45.2")
if is_transformers_version_greater_than_4_43():
import transformers.modeling_flash_attention_utils