THUDM/VisualGLM-6B
Chinese and English multimodal conversational language model | 多模态中英双语对话语言模型
PythonApache-2.0
Issues
- 0
请问作者或大家在训练自己的数据集的时候lora层的数量或者r应该怎么配置
#360 opened by zeroysss - 5
运行web_demo时一直无结果
#316 opened by hzh0720 - 0
- 0
- 2
微调时出现TypeError: string indices must be integers错误
#314 opened by 20191864218 - 3
AttributeError: 'ChatGLMTokenizer' object has no attribute 'sp_tokenizer'. Did you mean: '_tokenize'?
#323 opened by MetaDreamer3d - 0
关于纯文本微调
#357 opened by s2448587606 - 0
'Chatbot' object has no attribute 'style'
#356 opened by LFWOvO - 1
python web_demo.py报错
#352 opened by Dreamior - 0
finetune的时候加载模型失败
#355 opened by swaggy66 - 0
finetune的时候出现模型加载失败
#354 opened by swaggy66 - 0
运行web_demo_hf.py报错
#353 opened by floatingwh - 0
- 1
请问可以实现用qlora+model parallel 吗
#349 opened by zin-Fu - 2
- 0
qlora merge lora weights error
#350 opened by zousss - 1
Lora微调报错,fp16 is not supported
#348 opened by zousss - 0
关于只用文本数据集微调
#347 opened by SSXXXzjx - 5
visualglm进行QLoRA微调时报错,RuntimeError: mat1 and mat2 shapes cannot be multiplied (320x4096 and 1x25165824) [2024-03-07 07:26:17,037] [INFO] [launch.py:316:sigkill_handler] Killing subprocess 13210
#343 opened by munDane117 - 6
请问在阿里云上部署,连接不上huggingface网站的问题怎么解决呀?
#346 opened by 2232141528 - 2
微调之后加载web_demo时的报错
#345 opened by Venn2336 - 3
- 2
输入路径后出现None is not in list
#326 opened by 15235424244 - 0
请问怎么把finetune后的模型转成onnx格式呢?
#344 opened by ZCY311 - 7
No module named 'triton'
#332 opened by anonhl - 12
- 0
- 6
- 2
烦请帮忙看看,微调后运行cli_demo.py出现维度不一致问题; RuntimeError: The size of tensor a (12288) must match the size of tensor b (25165824) at non-singleton dimension 0
#339 opened by New-start-man - 0
运行finetune代码时,报错没有model_config.json
#341 opened by xiongxiaochu - 0
全代码开源会有吗
#340 opened by SamXiaosheng - 0
微调问题,微调后模型遗忘
#338 opened by sungatetop - 0
Lora微调返回代码-7
#337 opened by HarryK4673 - 1
Exception: Build 4bit layer failed. You need to install the latest bitsandbytes. Try `pip install bitsandbytes`. If you still meet error after installation, try running `from bitsandbytes.nn import LinearNF4` with python and fix the error.
#313 opened by LijunRio - 8
加载微调后模型时报错object of type 'QuantState' has no len()
#324 opened by KinokoY - 0
- 0
windows下微调
#334 opened by prettyprettyboy - 0
请问是否支持无图片的微调,如果支持,该如何修改代码?
#331 opened by chaijunmaomao - 1
deepspeed分布式训练出现sat ValueError inconsistent
#330 opened by elesun2018 - 0
huggingface模型要怎么微调,微调之后报错找不到xxxxxx.pt那个文件
#329 opened by dovs1314 - 8
AttributeError: module 'transformers_modules.hf.modeling_chatglm' has no attribute 'ChatGLMForConditionalGenerationWithImage'. Did you mean: 'ChatGLMForConditionalGeneration'?
#328 opened by w9931125 - 0
python main.py KeyError: 'type'
#327 opened by Mingzhu-Zhao - 4
您好,请问是否可以训练BLIP部分,微调代码似乎只可以微调chatglm的指定层的参数
#325 opened by chaijunmaomao - 0
ERROR: assert vocab_file is not None
#322 opened by fallinsilence - 4
- 0
关于微调之后的模型表现问题
#319 opened by 20191864218 - 3
python cli_demo.py 下载模型时候报错
#318 opened by xuyuhui666 - 1
什么时候开源基于GLM的中文CogVLM呢?
#317 opened by LutaoChu - 0
python api.py启动没报错,但是用客户端请求就报错
#312 opened by Rs529811807 - 4
直接源代码安装也是这个问题
#311 opened by Rs529811807