HaoZhongkai

icon indicating a website link https://haozhongkai.github.io/

icon company Tsinghua University icon location PhD student @ THU

Results 4 comments of HaoZhongkai

when running the official gsm8k with tool, multi turn async rollout sglang example without any modifications, the model crashes and appears Nan.

Hi, Did you resolve this problem? I also meet such problems when training with multi-turn sglang-async.

when running the official gsm8k with tool, multi turn async rollout sglang example without any modifications, the model crashes and appears Nan.

> > Hi, Did you resolve this problem? I also meet such problems when training with multi-turn sglang-async. > > [@HaoZhongkai](https://github.com/HaoZhongkai) > > Thanks to [@wuxibin89](https://github.com/wuxibin89) for his execellent work...

reinforce++运行之后卡住

请问你是卡在哪里了？是engine initialized process group那里吗

reinforce++运行之后卡住

那和我的报错不一样。我的通过 https://github.com/OpenRLHF/OpenRLHF/issues/492 解决了。 我在init engine后就开始训练了