423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-12-20 13:50:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix shift short attention

Browse Source 
Former-commit-id: 9a49cce8e6f6b222f74a07bdab40efee6a77b0f1
This commit is contained in:
hiyouga
2023-10-09 17:07:46 +08:00
parent b1bc191c45
commit 0c1e00574d
6 changed files with 46 additions and 52 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llmtuner/tuner/rm/workflow.py
View File

@@ -27,7 +27,7 @@ def run_rm(
dataset = get_dataset(model_args, data_args)
model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args, training_args.do_train, stage="rm")
dataset = preprocess_dataset(dataset, tokenizer, data_args, training_args, stage="rm")
data_collator = PairwiseDataCollatorWithPadding(tokenizer)
data_collator = PairwiseDataCollatorWithPadding(tokenizer, pad_to_multiple_of=4)
training_args_dict = training_args.to_dict()
training_args_dict.update(dict(remove_unused_columns=False)) # important for pairwise dataset