-
Notifications
You must be signed in to change notification settings - Fork 3.8k
Issues: hiyouga/LLaMA-Factory
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.
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
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
使用Qwen-7B使用Qlora时报错在阿里的PAI-DSW
pending
This problem is yet to be addressed
#5516
opened Sep 23, 2024 by
lzy728
1 task done
READ.ME中看到已经支持Qwen2.5(千问2.5)但是选择模版时,还是没有Qwen2和Qwen2.5的模版
pending
This problem is yet to be addressed
#5515
opened Sep 23, 2024 by
lishiyucn
1 task done
默认的optimizer是什么?如何添加自己的optimizer如SGD?
pending
This problem is yet to be addressed
#5514
opened Sep 23, 2024 by
DSW2001
1 task done
有可能对train函数加上差分隐私的训练处理吗,如果我想对sft微调训练过程中使用opacus加入差分隐私处理,我该怎么做?
pending
This problem is yet to be addressed
#5513
opened Sep 23, 2024 by
DSW2001
1 task done
How to train the mm_proj and the LLM part with lora of Qwen2-VL
pending
This problem is yet to be addressed
#5512
opened Sep 22, 2024 by
leoozy
1 task done
请问作者有计划支持序列并行相关的能力吗,类似于 xtuner 那种,类似于感觉可以集成 xtuner 的序列并行接口
pending
This problem is yet to be addressed
#5511
opened Sep 22, 2024 by
ldh127
1 task done
请问一下多图训练的时候如何指定每张图的像素?Internvl在训练的时候就有相关的功能
pending
This problem is yet to be addressed
#5509
opened Sep 22, 2024 by
leoozy
1 task done
Deepseek v2.5的 template 变了,与 v2不同
pending
This problem is yet to be addressed
#5506
opened Sep 21, 2024 by
piamo
1 task done
pretrain from scratch 输出都是数字
pending
This problem is yet to be addressed
#5504
opened Sep 21, 2024 by
UbeCc
1 task done
在checkpoint上继续训练,没有保存训练后的checkpint
pending
This problem is yet to be addressed
#5499
opened Sep 20, 2024 by
cuisws
1 task done
save_only_model后无法续训
pending
This problem is yet to be addressed
#5497
opened Sep 20, 2024 by
yuepengs
1 task done
Can you support Jamba 1.5 model and Mamba family models, mamba2-hybrid, ssm model, etc pls?
pending
This problem is yet to be addressed
#5496
opened Sep 20, 2024 by
badrabbitt
1 task done
用视频数据微调qwen2-vl-7b的算力要求是什么?
pending
This problem is yet to be addressed
#5493
opened Sep 20, 2024 by
J0eky
1 task done
昇腾910B npu8卡训练显存不足
npu
This problem is related to NPU devices
pending
This problem is yet to be addressed
#5491
opened Sep 20, 2024 by
LtroiNGU
1 task done
Is LLAVA chat template correct?
pending
This problem is yet to be addressed
#5489
opened Sep 20, 2024 by
mibejjh
1 task done
Running on machines with limited number of online programs
pending
This problem is yet to be addressed
#5488
opened Sep 19, 2024 by
moshushi007ow
1 task done
启动 webui失败
pending
This problem is yet to be addressed
#5485
opened Sep 19, 2024 by
ClementeGao
1 task done
请问DPO训练的时候有什么注意事项吗?我训练出来效果很差。
pending
This problem is yet to be addressed
#5484
opened Sep 19, 2024 by
zlh-source
1 task done
训练时template设为empty时,label开头会加上<|EOT|>,之前的版本好像不会这样
pending
This problem is yet to be addressed
#5474
opened Sep 18, 2024 by
haoranjun
只全参数微调Qwen2-VL-7B-Instruct的visual.merger部分,冻结其他模型参数,训练过程报错
pending
This problem is yet to be addressed
#5472
opened Sep 18, 2024 by
wjx-sudo
1 task done
sft do_predict, 生成的json 文件 的 label 都是空
pending
This problem is yet to be addressed
#5465
opened Sep 18, 2024 by
dayuyang1999
1 task done
AttributeError: 'Qwen2Attention' object has no attribute 'max_position_embeddings'
pending
This problem is yet to be addressed
#5461
opened Sep 17, 2024 by
chengchengpei
1 task done
Tips for implementing LlaMa-Factory for new Hardwares
pending
This problem is yet to be addressed
#5460
opened Sep 17, 2024 by
EtashGuha
no such a file or directory of data
pending
This problem is yet to be addressed
#5457
opened Sep 17, 2024 by
Esmail-ibraheem
1 task done
Previous Next
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.