LC1332/Chat-Haruhi-Suzumiya

Issues

• 使用vllm数据并行和ChatHaruhi一起使用会报RuntimeError: Cannot re-initialize CUDA in forked subprocess. To use CUDA with multiprocessing, you must use the 'spawn' start method

  #83 opened 2 months ago by 545771889a
  1

• 关于模型微调的一些问题

  #82 opened 3 months ago by HaoranZhangumich
  0

• 请教一下关于人格特质研究的MBTI部分的问题

  #81 opened 4 months ago by Darknessrky
  2

• 请问如何利用小说制作新的人物呢？

  #40 opened a year ago by zswodegit
  3

• silk-road / ChatHaruhi-RolePlaying这个数据集中的编码是什么？

  #79 opened 7 months ago by 545771889a
  4

• 请问训练大概1000条数据，需要什么样的配置？大概训练多久？

  #78 opened 8 months ago by samiytaa
  1

• 关于额外工具的一些问题？

  #25 opened a year ago by Hezhexi2002
  3

• 关于角色文本语料提取的系统

  #77 opened 8 months ago by new-xiao
  2

• 请问如何替换本地大语言模型

  #76 opened 9 months ago by smmmmall
  2

• 请问用基于Chat-Haruhi-Fusion_B微调的ChatGLM2的模型，为什么禁用采样了后，每次推理的结果都不一样

  #72 opened 10 months ago by renatz
  0

• 关于论文中提到的完整的s-R-D-H-q输入的具体实现

  #75 opened 9 months ago by GolfHotelSierra
  0

• 为什么不直接将提示词保存起来。下次直接使用呢

  #74 opened 10 months ago by steven-xiaver
  0

• 小说抽取还在继续吗？

  #73 opened 10 months ago by DeadLion
  0

• 请问角色扮演的实现是否仅仅基于数据集？

  #71 opened 10 months ago by CharlinChen
  1

• 关于Haruhi-Zero里角色的jsonl文件

  #70 opened a year ago by ShowSwichChen
  3

• Haruhi-Zero数据组织问题

  #69 opened a year ago by qinb
  4

• 关于几个训练细节的问题

  #63 opened a year ago by gctian
  5

• 关于角色构造的问题想问问您

  #66 opened a year ago by HelloNicoo
  4

• 关于训练的疑问

  #65 opened a year ago by cjw-d
  2

• 请问zero shot的能力如何？如果用户配置一个没有见过的role，可否通过一些ICL示例和proflie完成回答？

  #62 opened a year ago by Woooooody
  2

• About the character building of specific characters in the movie

  #60 opened a year ago by pipijiev12
  2

• Typo in README.md

  #61 opened a year ago by LazyGH
  2

• 你好！请问训练数据集方便公布嘛？

  #57 opened a year ago by LeiyanGithub
  3

• 我们在多轮对话的实践中发生了人格崩坏，希望得到回复

  #59 opened a year ago by Kirovsiki
  3

• 完成了ChatGLM3-base上的微调实践，但还是有一些问题

  #58 opened a year ago by zhaodice
  7

• Is there any support for customizing the persona to generate replies?

  #56 opened a year ago by iamlxb3
  1

• qwen1.8B的微调脚本链接跳转错误

  #55 opened a year ago by sanbuphy
  2

• 关于从小说中抽取出的角色语料生成可训练数据的步骤

  #53 opened a year ago by tinydust18
  7

• 关于训练和结果复现的疑问

  #51 opened a year ago by Zhuqln
  2

• 如何添加openai的代理网址呢?

  #43 opened a year ago by E-sion
  4

• 训练、数据问题请教

  #50 opened a year ago by goodnessSZW
  2

• 百川HF Demo

  #48 opened a year ago by MoidzzZ
  4

• 能否借鉴AI小镇的思路实现长期记忆？

  #45 opened a year ago by Winter-of-Cirno
  1

• 关于story的一些疑问

  #44 opened a year ago by Louis-y-nlp
  4

• Future Research Direction Exploring 未来研究方向探讨

  #33 opened a year ago by Neph0s
  6

• 多轮对话后对话陷入死循环

  #37 opened a year ago by pihang
  5

• ChatGLM2版的本地部署

  #41 opened a year ago by MoidzzZ
  4

• requirements.txt包含了冲突的依赖

  #39 opened a year ago by raven4752
  1

• novel_collecting项目的一些疑问

  #38 opened a year ago by Lvjinhong
  1

• 音频怎么加到界面上

  #31 opened a year ago by pihang
  5

• 关于训练数据使用问题

  #29 opened a year ago by lucasjinreal
  15

• ChatHaruhi-v1数据集 什么时候发布呀？

  #36 opened a year ago by Vincentyua
  0

• 可以将数据传huggingface吗？

  #27 opened a year ago by lucasjinreal
  3

• yuki_builder.py whisper 命令执行失败

  #23 opened a year ago by mabuyun
  2

• 小说数据如何转为对话数据

  #28 opened a year ago by lucasjinreal
  4

• 能添加更多電影或遊戲角色嗎?

  #26 opened a year ago by win10ogod
  1

• chat凉宫春日的CoT

  #24 opened a year ago by stevezxzhou
  5

• 微调模型和代码发布 chatglm2-lora model released

  #22 opened a year ago by rayvzn119
  3

• 论文有部分地方不理解，求解释

  #21 opened a year ago by stevezxzhou
  2

• Merge New Character

  #19 opened a year ago by LC1332
  0