Repository Issue Activity (beta)

#11004 图片视频数据混合训练时候dynamic_batching不完善导致死锁卡死

#11005 test

#11003 请教大佬，多轮工具训练过程中，max_prompt_length、max_response_length具体指什么

#10967 何时支持qwen3.5

#11000 GFPO's sample

#2 How to process data and save locally instead of push_to_hub?

#536 加载Geometry3k数据报错

#428 tensor_parallel_size设置为2时候报错

#10966 多轮数据格式

#505 [Dual Cards] Error while attaching to shared memory segment /dev/shm/nccl-Ќ (size 0), error: No such file or directory (2)

#667 Should I bind the GPU and the rank together?

#646 metadata is None

#655 无论如何配置，当前框架总是on policy吗？

#652 [docs] Add new awesome work using EasyR1

#649 AssertionError: Non-root FSDP instance's `_is_root` should not have been set yet or should have been set to `False`

#647 在训练时能否不适用wandb？

#644 Qwen3-VL video processor size problem

#641 训练报错 ValueError: Image features and image tokens do not match: tokens: 7536, features 9092

#637 ValueError: num_samples should be a positive integer value, but got num_samples=0

#639 基于自己的数据集报错

#10 Lora Support

#327 SFT+GRPO推理性能与训练过程中val精度不一致

#213 How to add neural reward model in EasyR1/examples/score_function ?

#630 Test Performance Impact: upgrade vllm to 0.10

#629 Test Performance Impact: fix multi modal data oom