FreedomIntelligence/GrammarGPT

Issues

• finetune.py中修改use_lora，训练时报错Expected a cuda device, but got: cpu

  #11 opened 8 months ago by lizhao-8202
  7

• 微调报错：modeling_utils.py中报错weight_size = weight.numel() * dtype_byte_size(weight.dtype) AttributeError: 'str' object has no attribute 'numel'

  #10 opened 8 months ago by lizhao-8202
  5

• 因为GPU显卡内存原因导致的python generate.py报错：torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 3.83 GiB. GPU

  #9 opened 8 months ago by lizhao-8202
  0

• pytorch_model-00001-of-00004.bin本地能用zip文件打开，执行generate.py报：RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory

  #8 opened 8 months ago by lizhao-8202
  6

• requirements.txt中git+https://github.com/huggingface/peft.git含义

  #6 opened 8 months ago by lizhao-8202
  3

• huggingface_hub.utils._validators.HFValidationError: Repo id must use alphanumeric chars or '-', '_', '.', '--' and '..' are forbidden, '-' and '.'

  #7 opened 8 months ago by lizhao-8202
  3

• 论文中数据增强的细节

  #5 opened 9 months ago by Daserxqc
  0

• 请问论文中提到的augmentation方法产生的数据包含在instruction.json文件中吗

  #4 opened a year ago by xray1111
  0

• OSError: Can't load tokenizer for './saved model'

  #3 opened a year ago by dedebao
  3

• 测试数据

  #2 opened a year ago by Jingsong-Yan
  1

• 关于推理

  #1 opened a year ago by huangChengz
  1