shibing624/textgen

Issues

• ChatGLM模型微调问题咨询

  #19 opened 2 years ago by alexhmyang
  34

• training_copyt5_model_demo.py

  #60 opened 3 months ago by ytWu1314
  3

• 部署问题

  #59 opened 4 months ago by yeyuan0620
  4

• 单机多卡训练方法

  #28 opened 7 months ago by hangzeli08
  4

• 大佬请问有没有针对chatglm3的评估脚本呀

  #58 opened 8 months ago by Franklin-L
  3

• 请问训练的数据集如何获取啊？

  #12 opened a year ago by p-moon
  5

• Bart长文本训练问题

  #14 opened 2 years ago by YoungChanYY
  5

• 读取数据问题

  #17 opened 2 years ago by MonkeyTB
  15

• 怎么指定单GPU

  #20 opened 2 years ago by MonkeyTB
  11

• 请问chatglm with lora 什么时候支持多卡fine tune啊

  #15 opened 2 years ago by Zarc98
  6

• 请教一下Chinese-Alpaca-Plus-13B继续做peft微调需要多少显存？

  #31 opened 2 years ago by bash99
  4

• 请问您的ptuning是V1吗，现在支持ptuning V2吗，不知道哪个在生成上好一点。您的几个微调方法lora，adalora，ptuning哪个效果好啊，full-finetuning一把学习率是多少啊～

  #30 opened 2 years ago by hangzeli08
  3

• 安装问题

  #34 opened 2 years ago by chengturbo
  2

• lora 训练参数

  #40 opened 2 years ago by MonkeyTB
  3

• 关于数据集处理

  #37 opened 2 years ago by kostya-sec
  4

• 多卡部署

  #39 opened 2 years ago by ZTurboX
  3

• hello master，BLOOM 模型 目前支持单机多卡训练么，如果支持是

  #42 opened 2 years ago by 116705792
  3

• how to use qlora ?

  #43 opened 2 years ago by feng-1985
  5

• 希望作者可以将最新的Aquila-7B和baichuan-7B模型集成进来

  #45 opened 2 years ago by AILWQ
  9

• torch.load()希望能够增加map_location

  #47 opened 2 years ago by PolarisRisingWar
  4

• 1. README文档的微信群和几个别的超链接打不开 2. 开始写使用教程（大概更新会比较慢）

  #48 opened 2 years ago by PolarisRisingWar
  1

• 是否支持ChatGLM2-6b的微调？

  #51 opened a year ago by svjack
  2

• 大佬什么时候支持一个chatRwkv, 这个现在用的人那么火，不搞一下

  #54 opened a year ago by gg22mm
  2

• gpt2训练学不到东西

  #56 opened a year ago by Lxhnnn
  2

• BART、T5和GPT2的模型输入问题

  #57 opened a year ago by YunweiDai
  0

• 大老看了您的微调方式，有点疑问，您的三个阶段是在哪里体现出来的？

  #55 opened a year ago by gg22mm
  2

• 模型结果有问题

  #53 opened a year ago by chw-shuai
  1

• bart纠错模型训练问题

  #49 opened a year ago by EASTERNTIGER
  2

• README中LLaMA模型部分是不是还没有更新+merge_peft_adapter.py的save_pretrained()似乎有问题

  #46 opened 2 years ago by PolarisRisingWar
  1

• ChatGLM-6B full parameter fine-tuning loss = 0

  #44 opened 2 years ago by xiaojunjun65
  1

• 模型训练出错

  #41 opened 2 years ago by svjack
  9

• NotImplementedError: Cannot copy out of meta tensor; no data!

  #36 opened 2 years ago by LMXKO
  3

• chatglm用lora训练完predict出的结果和重新加载模型和lora后输出的结果差异很大

  #27 opened 2 years ago by hongyix
  15

• T5、GPT用于文本翻译的句式在那个文件

  #11 opened 2 years ago by vikey
  2

• Bart文本生成

  #6 opened 2 years ago by suntea233
  3

• ImportError: cannot import name 'ChatGlmModel' from 'textgen'

  #32 opened 2 years ago by 888yyh
  1

• 保存模型问题

  #33 opened 2 years ago by MonkeyTB
  2

• 作者大哥，请问可以让chatglm系列支持单机多卡吗，改了一下午，改崩溃了，终于可以训练了，但是保存不了··· 你来吧

  #24 opened 2 years ago by hangzeli08
  2

• 用lora训练chatglm时跑内存

  #29 opened 2 years ago by NefelibataJay
  8

• 请问作者的学习率用的是多少啊，梯度下降方式呢，warmup步数呢，方便公开训练脚本参数吗

  #23 opened 2 years ago by hangzeli08
  1

• 请问有没有用于evaluation的代码？

  #26 opened 2 years ago by xiongxiaochu
  12

• 如何用使用 p40 训练lora

  #21 opened 2 years ago by hongyix
  2

• 数据集格式

  #5 opened 2 years ago by westwet
  3

• shibing624/prompt-t5-base-chinese测试结果和示例结果不一致

  #13 opened 2 years ago by hjing100
  2

• 如果我想训练一个细分领域的知识库，请问 训练数据 要按什么格式准备?您的例子给的是 纠正拼写错误，我们 知识库 训练需要改动哪些地方才能正确训练？

  #18 opened 2 years ago by alexhmyang
  1

• model 使用授权问题

  #22 opened 2 years ago by wzg-zhuo
  1

• Miss Requirement tensorboardX

  #16 opened 2 years ago by bash99
  8

• 您好，请教一下语法纠错后的PRF如何计算？

  #8 opened 2 years ago by ImXunan
  1

• 想要开启GPU上训练的话咋设置？

  #10 opened 2 years ago by huyi1989
  1

• 输入关键词 去生成关键词相关的文章用GPT2合适吗？

  #9 opened 2 years ago by huyi1989
  6