[BUG/Help] <max_sequence length>

[cxjtju]

Is there an existing issue for this?

• [X] I have searched the existing issues

Current Behavior

基于chatglm-6b进行增量预训练，请问训练数据每个样本的长度不能超过2048个token吗？

Expected Behavior

No response

Steps To Reproduce

长度超过2048会报ValueError: 130004 is not in list 参考这个repohttps://github.com/shibing624/MedicalGPT

Environment

- OS:
- Python:
- Transformers:
- PyTorch:
- CUDA Support (`python -c "import torch; print(torch.cuda.is_available())"`) :

Anything else?

No response