Haoran1234567

When I train train_sft with very small training data instructino_wild, it can run normally. When I increase some data, this happens, but the increased data is not much. i use...

Here is my run script： torchrun --standalone --nproc_per_node=8 train_sft.py \ --pretrain $PRETRAIN \ --model 'llama' \ --strategy colossalai_zero2 \ --log_interval 10 \ --save_path $SAVE_PATH \ --dataset $DATASET \ --batch_size 2...