HarderThenHarder/transformers_tasks

Issues

• 文本匹配任务Sentence Transformer疑问求解

  #90 opened 4 months ago by magicleo
  0

• 文本匹配任务的一些问题

  #89 opened 4 months ago by EvanSong77
  0

• ssl连接失败，无法下载 bert-base-chinese

  #88 opened 5 months ago by YKT-668
  0

• 大佬们看到help！！TypeError: tuple indices must be integers or slices, not str

  #87 opened 7 months ago by 11pomelo
  0

• 发现错误的代码

  #82 opened a year ago by Victoria-aaaaaa
  1

• UIE模型train文件运行报错

  #64 opened 2 years ago by White-bbb
  2

• 求解🙇🏻‍♀️标注平台打开网页失败 运行streamlit run playground_local.py --server.port 8001

  #86 opened a year ago by keyilizzie
  1

• 运行UIE train.sh时报错，求解答

  #85 opened a year ago by keyilizzie
  0

• 关于zero-shot

  #84 opened a year ago by pengcheng-yan
  0

• 关于ptuing

  #78 opened a year ago by init-neok
  4

• 关于UIE训练效果

  #41 opened 2 years ago by shenw8023
  1

• mbti_questions.json 这个测试文件可以分享一下吗

  #83 opened a year ago by zhongxing-mt
  1

• --mixed_precision=fp16

  #65 opened 2 years ago by WGS-note
  2

• where can I download checkpoints/model_1000？

  #70 opened 2 years ago by hsauod
  3

• 多卡训练后，推理时提示找不到tokenization_chatglm.py

  #69 opened 2 years ago by GromZhang
  5

• 多卡训练，无法运行train_multi_gpu.sh

  #52 opened 2 years ago by cheney369
  1

• 大佬您好，请问在做词表扩充的这个环节，为什么在llama的词表基础上扩充，而不是选择ChatGLM2的分词器

  #81 opened a year ago by GromZhang
  0

• 请问目前的对于chatglm的微调方法，对于新发布的chatglm2适用吗？

  #79 opened a year ago by GromZhang
  1

• 基于lora的方式finetune模型后，推理时间很长

  #80 opened a year ago by wangjiangyue0226
  0

• 请问，在您的zeroshot当中，仅仅是通过反馈给模型的history为我们设定的prompt，就完成了某一类的任务问题是吗？

  #67 opened 2 years ago by GromZhang
  1

• entity linking基于大模型的方案什么时候放出呢？有计划么

  #77 opened a year ago by amelieyu
  0

• 您这里的uie是生成式还是抽取式，cpu速度如何

  #75 opened a year ago by BEILOP
  0

• UIE抽取为什么加载不了模型

  #74 opened a year ago by Smile-shirley
  0

• PPO中critic模型不是应当使用reward model模型吗？

  #73 opened 2 years ago by zhangjian94cn
  0

• ValueError: Attempting to unscale FP16 gradients

  #45 opened 2 years ago by magnificent1208
  1

• UIE支持uie-m-large吗

  #72 opened 2 years ago by liaoweiguo
  0

• p-tuning 换成英文数据集会报错， raise ValueError(f'Lable Error: "{label}" not in label_dict {list(self.label_dict)}.')

  #71 opened 2 years ago by chengturbo
  0

• RuntimeError: CUDA out of memory. Tried to allocate 42.00 MiB (GPU 0; 22.02 GiB total capacity; 20.

  #60 opened 2 years ago by chenchaoac
  5

• 单卡lora微调需要多少显存？

  #63 opened 2 years ago by CCzzzzzzz
  4

• 训练时带参数--quantization_bit 4 报错RuntimeError: self and mat2 must have the same dtype

  #62 opened 2 years ago by shangzhensen
  3

• chatglm 微调

  #68 opened 2 years ago by xiaogogonuo
  0

• 请问数据格式为什么要是把instruct，input，answer这种放在context里面的格式，直接传统的instruct，input，output三元组不行吗，差异在哪里

  #61 opened 2 years ago by hangzeli08
  1

• 在finetune中，请问save_model的时候，是把glm6b的model和lora训练后的model合并成一个了是吗？

  #66 opened 2 years ago by GromZhang
  7

• LLM的finetune需要多大的显存啊？为什么我batch__size调到了1,24g的显存还是不够。。。。。。。

  #48 opened 2 years ago by SlothRan
  2

• 请问标注平台是否支持chatglm或者chatrwkv等其他的语言模型

  #59 opened 2 years ago by wzsage
  0

• 请问这个repo里的peft代码是做过什么定制化的改动吗？是否可以直接把peft安装到python环境里呢？

  #58 opened 2 years ago by zhoujz10
  0

• 关于多卡训练总iter数设置

  #56 opened 2 years ago by HXZhong1997
  1

• 关于UIE模型微调，如何进行一机多卡训练？

  #47 opened 2 years ago by zhougz2018
  2

• Hi,后面会考虑复现基于大模型如bloom、llama的rlhf流程代码？

  #57 opened 2 years ago by nieallen
  0

• 请问大模型微调需要多少显存？

  #51 opened 2 years ago by Crayon-s7
  3

• 考虑支持MOSS吗

  #55 opened 2 years ago by BestSongEver
  0

• 关于GLM finetune的OOM

  #54 opened 2 years ago by nuoma
  0

• 能去掉网上的关联吗，每次都要重新下载

  #44 opened 2 years ago by magnificent1208
  1

• PET损失函数代码

  #53 opened 2 years ago by haizeiwanglf
  0

• 模型多卡训练

  #50 opened 2 years ago by HocherCN
  0

• 关于词表问题

  #49 opened 2 years ago by magnificent1208
  0

• UI界面打开提示模型未被定义

  #43 opened 2 years ago by kbwzy
  6

• 大模型阅读理解微调

  #46 opened 2 years ago by tianchiguaixia
  0

• playground推理过慢

  #42 opened 2 years ago by nieallen
  2

• chatglm做finetune的时候报错

  #40 opened 2 years ago by tianmala
  0