-
Notifications
You must be signed in to change notification settings - Fork 3k
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
Qwen2 7B SFT 无法启动训练
pending
This problem is yet to be addressed.
#4220
opened Jun 11, 2024 by
rantianhua
1 task done
无法进行推理,可以微调以及加载模型。
pending
This problem is yet to be addressed.
#4218
opened Jun 11, 2024 by
GaoHZ1
1 task done
使用qwen7b对训练好的sft权重合并之后,进行chat,出现keyerror错误
pending
This problem is yet to be addressed.
#4211
opened Jun 11, 2024 by
cove1011
1 task done
0.8.1版本DeepSpeed 的 zero stage3报错
solved
This problem has been already solved.
#4209
opened Jun 11, 2024 by
xinyubai1209
1 task done
使用单机多卡微调Qwen2-72B
pending
This problem is yet to be addressed.
#4205
opened Jun 11, 2024 by
zjxxsr
1 task done
老师,metric.py中pred最终出现乱码怎么处理?
pending
This problem is yet to be addressed.
#4201
opened Jun 11, 2024 by
demouo
1 task done
llamafactory-cli train -h 初始化报错
pending
This problem is yet to be addressed.
#4200
opened Jun 11, 2024 by
970602
1 task done
Qwen2中间checkpoint执行generate不停
pending
This problem is yet to be addressed.
#4197
opened Jun 11, 2024 by
ssgg-code
1 task done
Tools section not added in custom dataset
pending
This problem is yet to be addressed.
#4187
opened Jun 10, 2024 by
hasan9090
1 task done
glm-4-9b-chat-1m do_predict得到的generated_predictions.jsonl中的label出现了\n和一些非数据集中的结果。
bug
Something isn't working
pending
This problem is yet to be addressed.
#4178
opened Jun 9, 2024 by
coasxu
1 task done
Do you have plans to add fine-tuning scripts for other multimodal large models? For example, Qwen_VL, LLaVA1.6, MiniGPT4, etc.
pending
This problem is yet to be addressed.
#4174
opened Jun 9, 2024 by
xdaiycl
1 task done
NPU glm-4-9b-chat API推理报错
pending
This problem is yet to be addressed.
#4165
opened Jun 8, 2024 by
msqp
1 task done
昇腾卡训练不支持offload
pending
This problem is yet to be addressed.
#4146
opened Jun 7, 2024 by
wangbing35
1 task done
关于Qwen2-72B 全量参数微调所需的显卡下限
pending
This problem is yet to be addressed.
#4141
opened Jun 7, 2024 by
zhangbin1997
1 task done
data.utils.split_dataset中的切分和随机逻辑能否迁移到data.loader.get_dataset中?
pending
This problem is yet to be addressed.
#4140
opened Jun 7, 2024 by
luoqishuai
1 task done
【NPU】GLM-4-9B-Chat PPO 出错
pending
This problem is yet to be addressed.
#4135
opened Jun 7, 2024 by
hunterhome
1 task done
qwen1.5_7B使用Zero-2方式在8张A100(40G)和64张910A(32G)上SFT,报OOM
pending
This problem is yet to be addressed.
#4133
opened Jun 7, 2024 by
xiaoruirui356
1 task done
Will you support HQQ quantization in the future?
enhancement
New feature or request
pending
This problem is yet to be addressed.
#4113
opened Jun 6, 2024 by
SJY8460
1 task done
sft+freeze训练internlm2-base-7b报错,RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn
pending
This problem is yet to be addressed.
#4101
opened Jun 6, 2024 by
1737686924
1 task done
Unable to run model.generate() for MoD model
pending
This problem is yet to be addressed.
#4063
opened Jun 4, 2024 by
Zkli-hub
1 task done
用openai库 请求时,流式请求时缺stream_options={"include_usage": True}的处理,用于计算流式tokens
pending
This problem is yet to be addressed.
#3998
opened May 30, 2024 by
sasicDHH
1 task done
Feature suggestion: cutoff_len could optionally drop too long examples from dataset.
pending
This problem is yet to be addressed.
#3995
opened May 30, 2024 by
s4s0l
How to specify eval set during training process?
pending
This problem is yet to be addressed.
#3974
opened May 30, 2024 by
may012345
1 task done
MODPO: Multi-Objective Direct Preference Optimization
enhancement
New feature or request
#3973
opened May 30, 2024 by
AlexYoung757
MoRA: High-Rank Updating for Parameter-Efficient Fine-Tuning
enhancement
New feature or request
pending
This problem is yet to be addressed.
#3970
opened May 29, 2024 by
backroom-coder
Previous Next
ProTip!
Find all open issues with in progress development work with linked:pr.