Skip to content

Issues: modelscope/ms-swift

GRPO (R1) 训练交流群
#3076 opened Feb 12, 2025 by Jintao-Huang
Open 3
ms-swift3 Suggestion Box
#2217 opened Oct 10, 2024 by Jintao-Huang
Open 35
SWIFT Support GRPO Agent Training
#3194 opened Feb 20, 2025 by tastelikefeet
Open 1
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

403 Open 1,370 Closed
403 Open 1,370 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

sample 不支持 external_plugins参数
#3333 opened Mar 1, 2025 by DogeWatch
NPU环境训练速度问题
#3331 opened Mar 1, 2025 by obj12
使用多机多卡微调qwen-2.5-vl模型报错：[ERROR:modelscope] The request model: Qwen/Qwen2___5-VL-3B-Instruct does not exist!
#3330 opened Mar 1, 2025 by ganchun1130
Why does it feel like the loss acc of the Qwen2.5-7B-Instruct model trained by lora has not changed much?
#3329 opened Mar 1, 2025 by safwaqf
时序特征平均池化在2.6以及3.1版本上的表现存在差异
#3328 opened Feb 28, 2025 by laserwave
qwen2.5 3b sft 多机训练 运行到固定的某个step会有一个节点显存 OOM
#3327 opened Feb 28, 2025 by phantooom
--attn_impl 设定为 'flash_attn' 时，evaluation报错padding side
#3325 opened Feb 28, 2025 by Chord-Chen-30
模型微调时提示No such file or directory: ':/usr/local/cuda/bin/nvcc
#3323 opened Feb 28, 2025 by yizhilu
qlora微调之后的模型为什么不能merge lora enhancement New feature or request
#3320 opened Feb 28, 2025 by ArlanCooper
internvl2_5微调出现几个问题，希望能帮一下忙
#3319 opened Feb 28, 2025 by zzyyxxznb
2
deepseekVL2 SFT微调后，推理报错
#3317 opened Feb 28, 2025 by wl654655902
2
最佳模型的checkpoint没被保存下来 bug Something isn't working
#3314 opened Feb 28, 2025 by joey9503
grpo训练如何制定训练集和测试集
#3313 opened Feb 28, 2025 by 0awei0
1
使用 eval评测时，如何看few-shot设置多少？
#3312 opened Feb 27, 2025 by user2311717757
1
【GRPO】【reward_model】how to use qwen7b as reward model to give reward?
#3310 opened Feb 27, 2025 by glennccc
4
Request for supporting REINFORCE++ algorithm enhancement New feature or request
#3309 opened Feb 27, 2025 by Tinyone
如何使用load_dataset进行数据切分
#3308 opened Feb 27, 2025 by ex-yanminmin001
【GRPO】multinode training cuda error
#3307 opened Feb 27, 2025 by glennccc
11
Pickling error: Mini-InternVL-Chat-2B-V1-5
#3303 opened Feb 27, 2025 by thekarthikprasad
2
GRPO Training Speed Testing good first issue Good for newcomers
#3302 opened Feb 27, 2025 by hjh0119
SFT后的模型无法进行vllm推理
#3295 opened Feb 26, 2025 by todayplusplus
2
欢迎各位加入多模态MLLM技术交流群！！
#3294 opened Feb 26, 2025 by yl-96
swift2.x与swift3.x使用相同的数据集得到的最终样本数不一致
#3292 opened Feb 26, 2025 by LuoYuanzhen
1
使用sft训练序列分类任务（实际上是图像分类任务），在推理时如何加载LoRA参数，让模型输出分类得分
#3291 opened Feb 26, 2025 by zxgravity
2
训练过程中卡死，进程处于睡眠状态，GPU利用率为0
#3290 opened Feb 26, 2025 by 2043078895
2
ProTip! Updated in the last three days: updated:>2025-02-26.