大模型-报错RuntimeError: Expected all tensors to be on the same device, but found at least two devices, c

报错

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:1 and cuda:0! (when checking argument for argument mat2 in method wrapper_CUDA_mm)

解决方案

docker run --gpus all -d -v 

改成

docker run --gpus "device=0" -d -v

在容器层面限制调用的GPU显卡，防止代码自动调用多张显卡，导致报错