zhou

xtuner在超过10000条数据集上运行正常，在1000条数据集上运行失败

1

xtuner训练llama3：8b时遇到下面错误，但是数据集扩大后就没问题，但是我现在需要在小数据集上训练。 ```10/08 19:59:50 - mmengine - INFO - xtuner_dataset_timeout = 1:00:00 Setting TOKENIZERS_PARALLELISM=false for forked processes. Traceback (most recent call last): File "/home/yihua/.conda/envs/xtuner/lib/python3.10/site-packages/xtuner/tools/train.py", line 360, in main() File "/home/yihua/.conda/envs/xtuner/lib/python3.10/site-packages/xtuner/tools/train.py", line...