liucongg/ChatGLM-Finetuning issues and pull requests

#148 - HELP！I changed the local model path, but I still have to go to Hugging Face when I run it

Issue - State: open - Opened by EdgeComputing0 about 2 months ago

#147 - 配置环境出错

Issue - State: closed - Opened by nothing7744 3 months ago

#146 - TypeError

Issue - State: open - Opened by ZCzzzzzz 3 months ago - 1 comment

#145 - P-tuning

Issue - State: open - Opened by JulyCaoJie 5 months ago

#144 - 支持glm4

Issue - State: open - Opened by tcxia 5 months ago

#143 - python3.8、3.9、3.10安装torch都报错

Issue - State: closed - Opened by ZzYAmbition 5 months ago - 1 comment

#142 - RuntimeError: Default process group has not been initialized, please make sure to call init_process_group.

Issue - State: open - Opened by cqray1990 6 months ago - 1 comment

#141 - 关于流水线并行的一个问题

Issue - State: open - Opened by Cheung-Z 6 months ago

#140 - 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).

Issue - State: open - Opened by ysz2000 8 months ago

#139 - 请问如何制作数据集

Issue - State: open - Opened by Franklin-L 8 months ago

#138 - 只有训练没有验证

Issue - State: open - Opened by NanZhang1991 9 months ago - 2 comments

#137 - 问题

Issue - State: open - Opened by wuguangshuo 10 months ago

#136 - 关于Pipeline Parallel的data_iter的问题

Issue - State: closed - Opened by Coobiw 10 months ago - 1 comment

#135 - 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.

Issue - State: open - Opened by dczhangii 10 months ago - 1 comment

#134 - ChatGLM3四卡训练出错了

Issue - State: open - Opened by eanfs 10 months ago - 1 comment

#133 - 有关混合精度训练的问题

Issue - State: open - Opened by zzhdbw 10 months ago

#132 - ptuning是损失为nan

Issue - State: open - Opened by silence-moon 10 months ago

#131 - chatglm3 单卡训练报错了

Issue - State: open - Opened by eanfs 10 months ago - 4 comments

#130 - 在RTX 4090 上微调chatglm3报这个错：Current loss scale already at minimum - cannot decrease scale anymore

Issue - State: open - Opened by 450586509 10 months ago - 1 comment

#129 - 使用README中的代码,报错TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

Issue - State: open - Opened by sunheyang1 10 months ago

#129 - 使用README中的代码,报错TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

Issue - State: open - Opened by sunheyang1 10 months ago

#128 - 没有验证过程?

Issue - State: open - Opened by xiaozhubenben 11 months ago

#128 - 没有验证过程?

Issue - State: open - Opened by xiaozhubenben 11 months ago

#127 - ChatGLM3训练时报错TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

Issue - State: open - Opened by AILWQ 11 months ago - 16 comments

#126 - --model_name_or_path', '../THUDM/chatglm-6b'参数什么错误

Issue - State: open - Opened by daiyucan 11 months ago - 1 comment

#126 - --model_name_or_path', '../THUDM/chatglm-6b'参数什么错误

Issue - State: open - Opened by daiyucan 11 months ago - 1 comment

#125 - RuntimeError: Error building extension 'fused_adam'

Issue - State: open - Opened by LeFuGang 11 months ago - 5 comments

#124 - python版本

Issue - State: closed - Opened by liujinchang 11 months ago - 1 comment

#124 - python版本

Issue - State: closed - Opened by liujinchang 11 months ago - 1 comment

#123 - peft==0.5.0怎么安装

Issue - State: closed - Opened by zhl970124 11 months ago - 1 comment

#123 - peft==0.5.0怎么安装

Issue - State: closed - Opened by zhl970124 11 months ago - 1 comment

#122 - 运行trainer时报错Error building extension 'fused_adam'

Issue - State: open - Opened by J-G-Y 11 months ago - 5 comments

#121 - 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'

Issue - State: open - Opened by zzdxjtu 11 months ago - 2 comments

#120 - 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

Issue - State: open - Opened by diorw 12 months ago - 1 comment

#120 - 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

Issue - State: open - Opened by diorw 12 months ago - 1 comment

#119 - 解决 AttributeError: 'ChatGLMTokenizer' object has no attribute 'tokenizer'

Issue - State: closed - Opened by gangtie95 12 months ago - 1 comment

#119 - 解决 AttributeError: 'ChatGLMTokenizer' object has no attribute 'tokenizer'

Issue - State: closed - Opened by gangtie95 12 months ago - 1 comment

#118 - 计划再支持下chatglm3不？

Issue - State: closed - Opened by zjcanjux 12 months ago - 2 comments

#118 - 计划再支持下chatglm3不？

Issue - State: closed - Opened by zjcanjux 12 months ago - 2 comments

#117 - 两张2080ti多卡加载微调

Issue - State: open - Opened by Lxhnnn about 1 year ago - 1 comment

#117 - 两张2080ti多卡加载微调

Issue - State: open - Opened by Lxhnnn about 1 year ago - 1 comment

#116 - OSError: ChatGLM2-6B is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models'

Issue - State: open - Opened by xxtyy about 1 year ago - 4 comments

#116 - OSError: ChatGLM2-6B is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models'

Issue - State: open - Opened by xxtyy about 1 year ago - 4 comments

#115 - 请问是否考虑支持其他开源llm呀？

Issue - State: closed - Opened by Irvingao about 1 year ago - 1 comment

#115 - 请问是否考虑支持其他开源llm呀？

Issue - State: closed - Opened by Irvingao about 1 year ago - 1 comment

#114 - mask_token = 150001

Issue - State: closed - Opened by Geministudents about 1 year ago - 1 comment

#114 - mask_token = 150001

Issue - State: closed - Opened by Geministudents about 1 year ago - 1 comment

#113 - 运行chatglm2单卡微调的时候遇到了ModuleNotFoundError: No module named 'peft'问题

Issue - State: closed - Opened by Alan-JW about 1 year ago - 3 comments

#113 - 运行chatglm2单卡微调的时候遇到了ModuleNotFoundError: No module named 'peft'问题

Issue - State: closed - Opened by Alan-JW about 1 year ago - 3 comments

#112 - tokenizer.bos_token_id 为啥不正确呢？

Issue - State: closed - Opened by sjyttkl about 1 year ago - 2 comments

#111 - 预测代码

Issue - State: closed - Opened by HongTu0319 about 1 year ago - 4 comments

#111 - 预测代码

Issue - State: closed - Opened by HongTu0319 about 1 year ago - 4 comments

#110 - 多轮对话代码

Issue - State: open - Opened by J-G-Y about 1 year ago - 1 comment

#110 - 多轮对话代码

Issue - State: open - Opened by J-G-Y about 1 year ago - 1 comment

#109 - predict error

Issue - State: open - Opened by tcxia about 1 year ago

#109 - predict error

Issue - State: open - Opened by tcxia about 1 year ago

#108 - 梯度叠加

Issue - State: closed - Opened by Cheung-Z about 1 year ago - 2 comments

#107 - 4090显卡有办法跑这个项目吗？

Issue - State: open - Opened by 2279072142 about 1 year ago - 3 comments

#106 - 求问，train_pipeline.py是否可以直接用在chatglm2上！

Issue - State: closed - Opened by Chtholly1 about 1 year ago - 1 comment

#105 - 预测代码

Issue - State: open - Opened by qaqrt about 1 year ago - 2 comments

#104 - 关于chatglm2与chatglm数据格式的问题

Issue - State: open - Opened by Kayce001 about 1 year ago - 2 comments

#104 - 关于chatglm2与chatglm数据格式的问题

Issue - State: open - Opened by Kayce001 about 1 year ago - 2 comments

#103 - 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

Issue - State: open - Opened by yangzhipeng1108 about 1 year ago - 1 comment

#102 - 执行 train.py过程报错 exits with return code = -9

Issue - State: open - Opened by yhx0105 about 1 year ago - 3 comments

#102 - 执行 train.py过程报错 exits with return code = -9

Issue - State: open - Opened by yhx0105 about 1 year ago - 3 comments

#101 - 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

Issue - State: open - Opened by karots123 about 1 year ago

#101 - 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

Issue - State: open - Opened by karots123 about 1 year ago

#100 - 单卡正常，多卡报错

Issue - State: open - Opened by SCU-JJkinging about 1 year ago - 5 comments

#100 - 单卡正常，多卡报错

Issue - State: open - Opened by SCU-JJkinging about 1 year ago - 5 comments

#99 - 增加chatGLM2+PT的预测代码

Pull Request - State: open - Opened by micrazy about 1 year ago - 9 comments

#98 - 请问是否支持断点续训？lora和全参微调

Issue - State: open - Opened by lianglinyi about 1 year ago - 1 comment

#98 - 请问是否支持断点续训？lora和全参微调

Issue - State: open - Opened by lianglinyi about 1 year ago - 1 comment

#97 - 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).

Issue - State: closed - Opened by yc930401 about 1 year ago - 2 comments

#97 - 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).

Issue - State: closed - Opened by yc930401 about 1 year ago - 3 comments

#96 - 模型预测问题

Issue - State: open - Opened by Linxuxin about 1 year ago - 6 comments

#96 - 模型预测问题

Issue - State: open - Opened by Linxuxin about 1 year ago - 6 comments

#95 - 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

Issue - State: open - Opened by smile-II about 1 year ago - 1 comment

#95 - 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

Issue - State: open - Opened by smile-II about 1 year ago - 1 comment

#94 - 大佬，请问这个工程能微调，通义千问，或者百川吗？

Issue - State: closed - Opened by HDHD-forever about 1 year ago - 1 comment

#94 - 大佬，请问这个工程能微调，通义千问，或者百川吗？

Issue - State: closed - Opened by HDHD-forever about 1 year ago - 1 comment

#93 - 预测代码都没得？

Issue - State: closed - Opened by Adam86546853 about 1 year ago - 3 comments

#92 - 关于Freeze的问题

Issue - State: open - Opened by vergil-ong over 1 year ago - 1 comment

#92 - 关于Freeze的问题

Issue - State: open - Opened by vergil-ong over 1 year ago - 1 comment

#91 - ImportError: /root/.cache/torch_extensions/py310_cu116/fused_adam/fused_adam.so: cannot open shared object file: No such file or directory

Issue - State: open - Opened by HDHD-forever over 1 year ago - 4 comments

#91 - ImportError: /root/.cache/torch_extensions/py310_cu116/fused_adam/fused_adam.so: cannot open shared object file: No such file or directory

Issue - State: open - Opened by HDHD-forever over 1 year ago - 4 comments

#90 - 使用master分支chatglm2，p-tuning微调后，如何预测啊，没有找到相应代码

Issue - State: open - Opened by LuoFanA595 over 1 year ago - 3 comments

#90 - 使用master分支chatglm2，p-tuning微调后，如何预测啊，没有找到相应代码

Issue - State: open - Opened by LuoFanA595 over 1 year ago - 3 comments

#89 - Is 1560 the minimum input sequence length for training?

Issue - State: open - Opened by Shuai-Xie over 1 year ago - 1 comment

#89 - Is 1560 the minimum input sequence length for training?

Issue - State: open - Opened by Shuai-Xie over 1 year ago - 1 comment

#88 - v0.1分支里面，运行预测文件predict_pt.py文件报错？有了解的吗？

Issue - State: open - Opened by qian2020 over 1 year ago - 1 comment

#88 - v0.1分支里面，运行预测文件predict_pt.py文件报错？有了解的吗？

Issue - State: open - Opened by qian2020 over 1 year ago - 1 comment

#87 - Is this line wrong?

Issue - State: open - Opened by iridescentee over 1 year ago

#87 - Is this line wrong?

Issue - State: open - Opened by iridescentee over 1 year ago

#86 - OSError: OSErrorOSErrorUnable to load weights from pytorch checkpoint file for '/ChatGLM2-6b/THUDM/chatglm2-6b/pytorch_model-00001-of-00007.bin' at 'ChatGLM2-6b/THUDM/chatglm2-6b/pytorch_model-00001-of-00007.bin'. If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True.

Issue - State: closed - Opened by byy-git over 1 year ago - 1 comment

#86 - OSError: OSErrorOSErrorUnable to load weights from pytorch checkpoint file for '/ChatGLM2-6b/THUDM/chatglm2-6b/pytorch_model-00001-of-00007.bin' at 'ChatGLM2-6b/THUDM/chatglm2-6b/pytorch_model-00001-of-00007.bin'. If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True.

GitHub / liucongg/ChatGLM-Finetuning issues and pull requests