From e82e6829da5aba623e66aa096e3e51137fdc2559 Mon Sep 17 00:00:00 2001 From: vvycaaa <147325516+vvycaaa@users.noreply.github.com> Date: Mon, 25 Dec 2023 20:28:10 +0800 Subject: [PATCH] bugfix: The 'preprocess_dataset' function call is missing required parameter (#192) Co-authored-by: github --- dbgpt_hub/train/sft_train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/dbgpt_hub/train/sft_train.py b/dbgpt_hub/train/sft_train.py index 2b7a31c..8dd2469 100644 --- a/dbgpt_hub/train/sft_train.py +++ b/dbgpt_hub/train/sft_train.py @@ -48,7 +48,7 @@ def run_sft( model, tokenizer = load_model_and_tokenizer( model_args, finetuning_args, training_args.do_train ) - dataset = preprocess_dataset(dataset, tokenizer, data_args, training_args) + dataset = preprocess_dataset(dataset, tokenizer, data_args, training_args, "sft") data_collator = DataCollatorForSeq2Seq( tokenizer=tokenizer, label_pad_token_id=IGNORE_INDEX