Skip to content

Issues: hiyouga/LLaMA-Factory

🚨FAQs | 常见问题🚨
#4614 opened Jun 28, 2024 by hiyouga
Open
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

200 Open 4,965 Closed
200 Open 4,965 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

单机多卡SFT，不同的rank训练步调不一致 solved This problem has been already solved
#5932 opened Nov 4, 2024 by kehuanfeng
1 task done
2
SFT后模型合并Lora权重，回答质量下降明显（非Issue #2505，#4913） pending This problem is yet to be addressed
#5930 opened Nov 4, 2024 by BGbigbear
1 task done
Use a LoRA finetuned model in Dify? pending This problem is yet to be addressed
#5928 opened Nov 4, 2024 by wingvortex
1 task done
ValueError:This model does not support image input. solved This problem has been already solved
#5918 opened Nov 3, 2024 by zjrwtx
1 task done
6
Hardware Requirement in the readme lacks critical info pending This problem is yet to be addressed
#5916 opened Nov 2, 2024 by xzuyn
llamafactory会考虑支持 Online DPO 吗 pending This problem is yet to be addressed
#5902 opened Nov 1, 2024 by piamo
1 task done
昇腾910b推理qwen2-7b npu This problem is related to NPU devices pending This problem is yet to be addressed
#5894 opened Nov 1, 2024 by leoneyar
1 task done
如何添加额外的可训练参数？ pending This problem is yet to be addressed
#5891 opened Nov 1, 2024 by Zheng-Jay
1 task done
与LLaVA官方代码训练结果性能相差较大 pending This problem is yet to be addressed
#5890 opened Nov 1, 2024 by zhipeixu
1 task done
[trainer_utils.py] Why layerwise GaLoRE optimizer does not support gradient accumulation, any underlining reasons? pending This problem is yet to be addressed
#5887 opened Nov 1, 2024 by oncleJules
1 task done
How to mask out specific chunks for loss calculation pending This problem is yet to be addressed
#5886 opened Oct 31, 2024 by Hanzhang-lang
1 task done
求助！ npu This problem is related to NPU devices pending This problem is yet to be addressed
#5884 opened Oct 31, 2024 by lucky0223
显存充足，无法调用，显示只使用一点显存 pending This problem is yet to be addressed
#5878 opened Oct 30, 2024 by Lgugeng
1 task done
1
openapi.json 没有上传文件相关的接口,怎么实现api推荐分析文件啊,这样多模态才能调试 pending This problem is yet to be addressed
#5876 opened Oct 30, 2024 by a67793581
1 task done
1
微调qwen2.5 3B模型报“UnicodeDecodeError”错误，请作者帮忙看看，谢谢！ pending This problem is yet to be addressed
#5875 opened Oct 30, 2024 by yangdy11111
1 task done
对qwen2.5-14B增量预训练后推理时，部分重复一段话 pending This problem is yet to be addressed
#5872 opened Oct 30, 2024 by Ayanami07
1 task done
4
视频使用mkv文件报错 pending This problem is yet to be addressed
#5870 opened Oct 30, 2024 by HelloWorld506
1 task done
请问一下 什么时候支持openbmb/MiniCPM-V-2_6 这个多模态的微调 谢谢 pending This problem is yet to be addressed
#5869 opened Oct 30, 2024 by ML-GCN
1 task done
template formatter能否支持一定程度上的逻辑判断？ pending This problem is yet to be addressed
#5868 opened Oct 30, 2024 by Ricardo-L-C
1 task done
Question regarding Function Calling in ShareGPT format pending This problem is yet to be addressed
#5866 opened Oct 30, 2024 by emrecanacikgoz
1 task done
When exporting, drop unused parameters instead of erroring pending This problem is yet to be addressed
#5853 opened Oct 29, 2024 by inflatebot
1 task done
Newcomer for help: If the same training corpus is used, is there a way to save the pre-tokenized data and load it directly next time? pending This problem is yet to be addressed
#5851 opened Oct 29, 2024 by Wiselnn570
1 task done
1
在A40 96G显存上对llama-3.1-70B-instruction通过QLoRA微调成功也导出成功，想在只有CPU的服务器上运行，提示You are trying to offload the whole model to the disk. Please use the disk_offload function instead pending This problem is yet to be addressed
#5849 opened Oct 28, 2024 by gannyee
1 task done
How to continue training LoRA made without llama factory? pending This problem is yet to be addressed
#5848 opened Oct 28, 2024 by Sehyo
1 task done
2
Support ferretui model pending This problem is yet to be addressed
#5847 opened Oct 28, 2024 by dushwe
ProTip! Exclude everything labeled bug with -label:bug.