liucongg / ChatGLM-Finetuning

基于ChatGLM-6B、ChatGLM2-6B、ChatGLM3-6B模型，进行下游具体任务微调，涉及Freeze、Lora、P-tuning、全参微调等

liucongg/ChatGLM-Finetuning Issues

4090显卡有办法跑这个项目吗？
Updated 2 months ago3
RuntimeError: The server socket has failed to listen on any local network address. The server socket has failed to bind to [::]:520 (errno: 13 - Permission denied). The server socket has failed to bind to ?UNKNOWN? (errno: 13 - Permission denied).
Updated 2 months ago
请问如何制作数据集
Updated 2 months ago
在RTX 4090 上微调chatglm3报这个错：Current loss scale already at minimum - cannot decrease scale anymore
Updated 2 months ago1
ChatGLM3四卡训练出错了
Updated 2 months ago1
tokenizer.bos_token_id 为啥不正确呢？
Closed 5 months ago2
只有训练没有验证
Updated 2 months ago1
RuntimeError: Error building extension 'fused_adam'
Updated 2 months ago2
File "/root/anaconda3/envs/llm/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 847, in __init__ setattr(self, key, value) AttributeError: can't set attribute 'eos_token'
Updated 3 months ago2
chatglm3 单卡训练报错了
Updated 3 months ago4
问题
Updated 4 months ago
关于Pipeline Parallel的data_iter的问题
Closed 4 months ago1
local_rank默认设置-1，在执行torch.distributed.get_rank()时报错：RuntimeError: Default process group has not been initialized, please make sure to call init_process_grroup.
Updated 4 months ago
有关混合精度训练的问题
Updated 4 months ago
ptuning是损失为nan
Updated 4 months ago
预测代码
Updated 4 months ago2
使用README中的代码,报错TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType
Updated 4 months ago
没有验证过程?
Updated 5 months ago
ChatGLM3训练时报错TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType
Updated 5 months ago14
计划再支持下chatglm3不？
Closed 5 months ago2
解决 AttributeError: 'ChatGLMTokenizer' object has no attribute 'tokenizer'
Closed 5 months ago1
运行trainer时报错Error building extension 'fused_adam'
Updated 5 months ago4
python版本
Closed 5 months ago1
peft==0.5.0怎么安装
Closed 5 months ago1
关于chatglm2与chatglm数据格式的问题
Updated 5 months ago2
ChatGLM3-6B-32k 请问chatglm3里面的代码除了模型名字以外别的需要修改吗
Updated 5 months ago1
多机多卡运行 stage2 和stage3，stage3的训练时间是stage2的25倍，这结果合理吗
Updated 5 months ago1
训练显卡用不了Unable to determine the device handle for GPU0000:05:00.0: Unknown Error 需要重启显卡
Updated 5 months ago1
关于Freeze的问题
Updated 5 months ago1
--model_name_or_path', '../THUDM/chatglm-6b'参数什么错误
Updated 5 months ago1
请问是否考虑支持其他开源llm呀？
Closed 5 months ago1
求问，train_pipeline.py是否可以直接用在chatglm2上！
Closed 5 months ago1
梯度叠加
Closed 5 months ago2
预测代码都没得？
Closed 5 months ago3
大佬，请问这个工程能微调，通义千问，或者百川吗？
Closed 5 months ago1
The server socket has failed to listen on any local network address. The server socket has failed to bind to [::]:520 (errno: 13 - Permission denied).
Closed 9 months ago2
单卡正常，多卡报错
Updated 5 months ago5
预测代码
Closed 5 months ago4
两张2080ti多卡加载微调
Updated 6 months ago1
ImportError: /root/.cache/torch_extensions/py310_cu116/fused_adam/fused_adam.so: cannot open shared object file: No such file or directory
Updated 6 months ago4
OSError: ChatGLM2-6B is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models'
Updated 6 months ago4
mask_token = 150001
Closed 7 months ago1
使用master分支chatglm2，p-tuning微调后，如何预测啊，没有找到相应代码
Updated 6 months ago3
多轮对话代码
Updated 7 months ago1
运行chatglm2单卡微调的时候遇到了ModuleNotFoundError: No module named 'peft'问题
Closed 7 months ago3
predict error
Updated 7 months ago
模型预测问题
Updated 7 months ago6
执行 train.py过程报错 exits with return code = -9
Updated 8 months ago3
关于 RuntimeError: element 0 or 1 of tensors does not require grad and does not have a grad_fn的问题讨论
Updated 9 months ago
请问是否支持断点续训？lora和全参微调
Updated 9 months ago1