在加载预训练模型进行训练时,在optimizer.step()行会出现
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0
因为加载预训练模型时,也加载优化器,如下所示
optimizer.load_state_dict(torch.load(optimizer_path))
可以尝试把模型先移动到GPU上,然后在加载optimizer.load_state_dict(...)