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

344 Open 5,786 Closed
344 Open 5,786 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

GPU Imbalanced Loading bug Something isn't working pending This problem is yet to be addressed
#7250 opened Mar 11, 2025 by WillDreamer
1 task done
1
微调 DeepSeek-R1 蒸馏模型，在 Chat 加载秩表现出色，但在导出部署到 Ollama 后问答准确率大幅下降 bug Something isn't working pending This problem is yet to be addressed
#7238 opened Mar 11, 2025 by Nehcknarf
1 task done
2
单机多卡（4 x 3090）Linux 系统 使用默认的llamafactory-cli train /homeqwen3b_lora_pretrain.yaml 报错 bug Something isn't working pending This problem is yet to be addressed
#7233 opened Mar 10, 2025 by Johnnythefool
1 task done
3
Error when training bug Something isn't working pending This problem is yet to be addressed
#7232 opened Mar 10, 2025 by catn1pdeal3r
1 task done
1
安装报错：Failed to build autoawq==0.2.8 bug Something isn't working pending This problem is yet to be addressed
#7225 opened Mar 9, 2025 by thinkingInWorldByNull
1 task done
1
希望提供对phi4-mini:3.8b的支持。 enhancement New feature or request pending This problem is yet to be addressed
#7224 opened Mar 9, 2025 by liuaifu
1 task done
1
raise RuntimeError("Cannot find valid samples, check data/README.md for the data format.") when wikipedia_en bug Something isn't working pending This problem is yet to be addressed
#7220 opened Mar 8, 2025 by new-Sunset-shimmer
1 task done
1
vllm_infer对qwen2.5vl推理很慢，10000个图文对卡住很久 bug Something isn't working pending This problem is yet to be addressed
#7216 opened Mar 8, 2025 by 2019211753
1 task done
TypeError: unhashable type: 'list' bug Something isn't working pending This problem is yet to be addressed
#7214 opened Mar 7, 2025 by CaiJichang212
1 task done
1
Reward Model 推理 bug Something isn't working pending This problem is yet to be addressed
#7212 opened Mar 7, 2025 by SFTJBD
1 task done
训练deepseek蒸馏的7B时，loss在每个epoch开始时翻倍 bug Something isn't working pending This problem is yet to be addressed
#7208 opened Mar 7, 2025 by Y56611
1 task done
1
同一个数据集和模型，相同参数设置，训练两次，0.5epoch时会因为模型见到数据顺序不同的原因导致很大效果差异吗？ invalid This doesn't seem right
#7200 opened Mar 7, 2025 by tiphaineeee
1 task done
when will you release the new version? bug Something isn't working pending This problem is yet to be addressed
#7199 opened Mar 7, 2025 by ganisback
1 task done
deepseek r1 微调后我应该怎么加载lora参数推理呢 bug Something isn't working pending This problem is yet to be addressed
#7185 opened Mar 6, 2025 by joyyyhuang
1 task done
3
使用unsloth加速报错 bug Something isn't working pending This problem is yet to be addressed
#7177 opened Mar 6, 2025 by GEK1
1 task done
5
MiniCPM-o-2_6的sft、lora训练报错：Some weights of the model checkpoint at /app123/model/MiniCPM-o-2_6 were not used when initializing MiniCPMO: bug Something isn't working pending This problem is yet to be addressed
#7169 opened Mar 5, 2025 by winni0
1 task done
deepseek-moe-16B预训练问题 bug Something isn't working pending This problem is yet to be addressed
#7165 opened Mar 5, 2025 by zyp-byte
1 task done
2
跑open_r1_math数据集，qwen7b-instruct每次跑到53个step报错 bug Something isn't working pending This problem is yet to be addressed
#7163 opened Mar 5, 2025 by fsq77
1 task done
2
Qwen/Qwen2.5-VL-7B-Instruct PPO 训练报错 bug Something isn't working pending This problem is yet to be addressed
#7159 opened Mar 5, 2025 by ulovecode
1 task done
2
qwen2.5vl 開啟unsloth時，使用lora检查点繼續訓練時出錯。 bug Something isn't working pending This problem is yet to be addressed
#7156 opened Mar 4, 2025 by mpeilun
1 task done
Errors when directly calling the "run_exp()" function under the "train" command bug Something isn't working pending This problem is yet to be addressed
#7155 opened Mar 4, 2025 by Soever
1 task
单机单卡SFT比单机多卡deepspeed Zero3效果好？？？ bug Something isn't working pending This problem is yet to be addressed
#7153 opened Mar 4, 2025 by Essence9999
1 task done
webui上选择的是bf16, 跑的时候报错并提示只支持bf16 bug Something isn't working pending This problem is yet to be addressed
#7151 opened Mar 4, 2025 by xudong2019
1 task done
After updating the version, I attempted to train qwen2_vl but encountered issues with slower training speed and decreased accuracy. I have not been able to identify the cause. enhancement New feature or request pending This problem is yet to be addressed
#7150 opened Mar 4, 2025 by xueaa
1 task done
4
OSError: [Errno 7] Argument list too long bug Something isn't working pending This problem is yet to be addressed
#7144 opened Mar 3, 2025 by leoozy
1 task done
1
ProTip! Find all open issues with in progress development work with linked:pr.