'ModelArguments' object has no attribute 'use_v2'

[Glen-Chen-Blue]

╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮ │ /home/wmnlab/chat/ChatGLM-Efficient-Tuning/src/train_bash.py:21 in │ │ │ │ 18 │ │ 19 │ │ 20 if name == "main": │ │ ❱ 21 │ main() │ │ 22 │ │ │ │ /home/wmnlab/chat/ChatGLM-Efficient-Tuning/src/train_bash.py:8 in main │ │ │ │ 5 │ model_args, data_args, training_args, finetuning_args, general_args = get_train_args │ │ 6 │ │ │ 7 │ if general_args.stage == "sft": │ │ ❱ 8 │ │ run_sft(model_args, data_args, training_args, finetuning_args) │ │ 9 │ elif general_args.stage == "rm": │ │ 10 │ │ run_rm(model_args, data_args, training_args, finetuning_args) │ │ 11 │ elif general_args.stage == "ppo": │ │ │ │ /home/wmnlab/chat/ChatGLM-Efficient-Tuning/src/glmtuner/tuner/sft/workflow.py:30 in run_sft │ │ │ │ 27 │ │ tokenizer=tokenizer, │ │ 28 │ │ model=model, │ │ 29 │ │ ignore_pad_token_for_loss=(data_args.ignore_pad_token_for_loss and not training_ │ │ ❱ 30 │ │ use_v2=model_args.use_v2 │ │ 31 │ ) │ │ 32 │ │ │ 33 │ # Override the decoding parameters of Seq2SeqTrainer │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────╯ AttributeError: 'ModelArguments' object has no attribute 'use_v2'

[canglincuizhu]

The code has been updated just now. And 'use_v2' was removed.

[hiyouga]

Please update your code.