THUDM/AgentTuning

Issues

• 训练数据中指令与模型行为不匹配

  #58 opened 2 months ago by haichao592
  0

• 本地模型

  #57 opened 3 months ago by lz2021211161
  0

• 请问哪里可以找到工作里对于数据库方面的训练数据

  #56 opened 3 months ago by Mucalinda2436
  1

• weight decay确定是0.1吗？

  #54 opened 4 months ago by Fu-Dayuan
  1

• 魔塔上的 AgentInstruct 数据集的 conversation 都是空值

  #55 opened 4 months ago by XianglongTan
  0

• 基于fastchat部署，推理异常

  #46 opened 6 months ago by ruifengma
  3

• 貌似hotpotqa测试脚本跑不起来？

  #53 opened 5 months ago by Fu-Dayuan
  1

• 训练数据是如何采样的？

  #52 opened 5 months ago by Fu-Dayuan
  3

• 通用数据如何筛选

  #41 opened 7 months ago by LuoKaiGSW
  7

• Can you point to the ShareGPT filtered/cleaned data used?

  #50 opened 5 months ago by harshraj172
  1

• if it is possible to conduct RLHF from env

  #51 opened 5 months ago by SHITIANYU-hue
  1

• Can I run AgentInstruct data on the AgentBench?

  #49 opened 5 months ago by harshraj172
  1

• 可以给个简单点的工具调用示例吗

  #48 opened 6 months ago by qq594495953
  1

• 期待用 Qwen72B 训练的模型。

  #47 opened 6 months ago by milomoon
  1

• 除了用docker运行，还有其他方式可以运行AgentLM吗？

  #42 opened 6 months ago by caizhuoyue77
  6

• 关于TRAJECTORY FILTERING问题

  #44 opened 7 months ago by QingChengLineOne
  3

• AgentTuning 7b evaluate in HH， not expect as paper result

  #39 opened 7 months ago by Dhaizei
  13

• 请问下agentlm-7b最少需要多少显存可以推理

  #45 opened 7 months ago by nicolasNi
  5

• 论文中关于损失函数E的问题

  #10 opened 7 months ago by Dhaizei
  7

• Adding Contributors Section in readme.md file.

  #27 opened 7 months ago by mohitd404
  0

• Finetuning with Mistral or Yi?

  #43 opened 7 months ago by jFkd1
  1

• 关于数据集

  #37 opened 7 months ago by DryPilgrim
  0

• Number of training steps

  #36 opened 7 months ago by Mayer123
  1

• Dataset details 中找不到reward的计算方式

  #40 opened 7 months ago by DryPilgrim
  5

• 关于dataset statics 和 download

  #38 opened 7 months ago by DryPilgrim
  3

• 关于reward

  #32 opened 7 months ago by DryPilgrim
  2

• agent tuning和toolbench的区别

  #34 opened 7 months ago by Connor-Shen
  1

• 微调显存

  #35 opened 7 months ago by Reason-Wang
  1

• Start TGI worker

  #33 opened 7 months ago by mayilin0714
  1

• 请教reward分数的各种情况

  #30 opened 7 months ago by DryPilgrim
  1

• requests.exceptions.MissingSchema: Invalid URL '127.0.0.123332/generate': No scheme supplied. Perhaps you meant https://127.0.0.123332/generate?

  #31 opened 7 months ago by mayilin0714
  1

• Inference with `vllm`

  #29 opened 7 months ago by yc1999
  1

• 论文中Table 2中的数字的含义和计算方式

  #17 opened 7 months ago by DryPilgrim
  2

• Add license

  #24 opened 7 months ago by dalvishruti14
  0

• Grammer mistake in readme

  #19 opened 7 months ago by shraddha761
  1

• Auto comment

  #21 opened 7 months ago by shraddha761
  1

• 什么时候上魔塔社区

  #18 opened 7 months ago by QingChengLineOne
  1

• 论文中的问题

  #16 opened 7 months ago by QingChengLineOne
  1

• Model Output Length

  #11 opened 8 months ago by HeimingX
  2

• 是否可以不在docker里运行

  #5 opened 8 months ago by QingChengLineOne
  2

• 底座模型基于llama2，是否支持中文呢

  #4 opened 8 months ago by cnsky2016
  1

• fine-tune code

  #12 opened 8 months ago by faaany
  1

• 微调

  #9 opened 8 months ago by Dhaizei
  2

• An open queston: What's the difference between Agents and Tools

  #7 opened 8 months ago by zhaochenyang20
  0

• 交互轨迹的Reward如何得到

  #8 opened 8 months ago by jiezhangGt
  1

• AgentLM能支持openai.api类的接口本地部署吗？

  #3 opened 8 months ago by HuntZhaozq
  1

• train code

  #1 opened 8 months ago by wangyanli3630
  3