Traceback (most recent call last):
File "train.py", line 199, in
cost = trainBatch(crnn, criterion, optimizer)
File "train.py", line 186, in trainBatch
cost.backward()
File "/home/zhangmingzhou1/anaconda3/envs/torch/lib/python3.6/site-packages/torch/tensor.py", line 195, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File "/home/zhangmingzhou1/anaconda3/envs/torch/lib/python3.6/site-packages/torch/autograd/init.py", line 99, in backward
allow_unreachable=True) # allow_unreachable flag
File "/home/zhangmingzhou1/anaconda3/envs/torch/lib/python3.6/site-packages/torch/autograd/function.py", line 77, in apply
return self._forward_cls.backward(self, *args)
File "/home/zhangmingzhou1/anaconda3/envs/torch/lib/python3.6/site-packages/warpctc_pytorch-0.1-py3.6-linux-x86_64.egg/warpctc_pytorch/init.py", line 51, in backward
_grad_output = grad_output.to(ctx.grads.device)
RuntimeError: CUDA error: an illegal memory access was encountered
我训练时也出现了类似的错误,请问你这个错误解决了吗。
Try decreasing batch size. Sources: pytorch/pytorch#46974 and pytorch/pytorch#21819
Try decreasing batch size. Sources: pytorch/pytorch#46974 and pytorch/pytorch#21819
my batchsize is one,but failed