PyTorch 错误 RuntimeError: CUDA error: device-side assert triggered

最新推荐文章于 2025-03-25 09:44:32 发布

原创最新推荐文章于 2025-03-25 09:44:32 发布 · 7.2w 阅读

17 ·

本内容遵循CC 4.0 BY-SA版权协议

收录于

PyTorch

本文解析了PyTorch中CUDA错误：device-side assert触发的问题，详细介绍了错误发生的具体位置及原因，并给出了调整模型输出或标签类别数量以解决类别不匹配问题的方法。

PyTorch 错误 RuntimeError: CUDA error: device-side assert triggered

flyfish
具体提示如下

/pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:108: cunn_ClassNLLCriterion_updateOutput_kernel: block: [0,0,0], thread: [2,0,0] Assertion `t >= 0 && t < n_classes` failed.
/pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:108: cunn_ClassNLLCriterion_updateOutput_kernel: block: [0,0,0], thread: [20,0,0] Assertion `t >= 0 && t < n_classes` failed.
Traceback (most recent call last):
  File "train_feature.py", line 65, in run
    loss.backward()
  File "/media/ubuntu/data/pytorch1.7/lib/python3.6/site-packages/torch/tensor.py", line 221, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph)
  File "/media/ubuntu/data/pytorch1.7/lib/python3.6/site-packages/torch/autograd/__init__.py", line 126, in backward
    grad_tensors_ = _make_grads(tensors, grad_tensors_)
  File "/media/ubuntu/data/pytorch1.7/lib/python3.6/site-packages/torch/autograd/__init__.py", line 51, in _make_grads
    new_grads.append(torch.ones_like(out, memory_format=torch.preserve_format))
RuntimeError: CUDA error: device-side assert triggered