step2 without any response for a long time

[asfadfaf]

Creating prompt dataset ['/mnt/workspace/RLHF/data/pinpai'], reload=False
Creating dataset data_pinpai for train_phase=2 size=392
Creating dataset data_pinpai for train_phase=2 size=392
Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root...

I am training a reward model on my own dataset, and it gets stuck after loading the content mentioned above, without any response for a long time. Previously, I successfully ran it on a public dataset, but now it also gets stuck at step 2 on the public dataset. The training for step 1 on my own dataset was successfully completed. Why is this happening?