RuntimeError: unscale_() has already been called on this optimizer since the last update

[monkeyshichi]

运行命令如下： CUDA_VISIBLE_DEVICES=1 python src/train_pt.py
--model_name_or_path baichuan-7B
--do_train
--lora_target W_pack
--dataset wiki_demo
--finetuning_type lora
--quantization_bit 4
--output_dir pretrained
--overwrite_cache
--per_device_train_batch_size 4
--gradient_accumulation_steps 4
--lr_scheduler_type cosine
--logging_steps 10
--save_steps 1000
--learning_rate 5e-5
--num_train_epochs 3.0
--plot_loss
--fp16