使用Deepseek-lite训练DPO，显示expected mat1 and mat2 to have the same type, but got: float != c10: : BFLoat16

OpenLLMAI / OpenRLHF

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & Mixtral)

https://openrlhf.readthedocs.io/

Apache License 2.0

1.71k stars 160 forks source link

Open victorShawFan opened 1 month ago

victorShawFan commented 1 month ago

hijkzzz commented 1 month ago

通常这种情况可以尝试升级 deepspeed 到最新版（但是也不能确定 100% work）

victorShawFan commented 1 month ago

确实没有work

lixsh6 commented 4 weeks ago

确实没有work

请问有fix的办法吗？同遇到类似问题