All issues

Search results

grpo训练问题

Status: Open.

#10In QunBB/DeepLearning;

· skylarzhao1 opened on Aug 23, 2025

为什么我在执行sft_train的时候报错：RuntimeError: The size of tensor a (8) must match the size of tensor b (16) at non-singleton dimension 0

Status: Open.

#8In QunBB/DeepLearning;

· pengjunj opened on Apr 9, 2025

grpo_train的过程中，打印了模型推理结果，其中的Extracted是什么，为什么会有它？

Status: Open.

#7In QunBB/DeepLearning;

· winni0 opened on Mar 20, 2025

执行grpo_train的时候，报错：Expected all tensors to be on the same device, but found at least two devices

Status: Open.

#6In QunBB/DeepLearning;

· winni0 opened on Mar 19, 2025

deepseek-train项目里面环境问题

Status: Open.

#4In QunBB/DeepLearning;

· winni0 opened on Mar 17, 2025

环境配置

Status: Open.

#3In QunBB/DeepLearning;

· cpy18727 opened on Feb 22, 2025