英文:
Why conv2d yields different results with different batch size
问题
我使用相同的数据但不同的批次大小(使用stack)作为输入来提供conv2d:
a = torch.rand(1, 512, 16, 16) # (1, 512, 16, 16)
b = torch.cat([a, a, a], dim=0) # (3, 512, 16, 16)
a, b = a.cuda(), b.cuda()
net = nn.Conv2d(512, 1024, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
net = net.cuda()
ay = net(a)
by = net(b)
print('ay[0], by[0] max diff', torch.max(torch.abs(ay[0] - by[0])).item())
print('ay[0], by[0] allclose', torch.allclose(ay[0], by[0]))
然而,结果不同:
ay[0], by[0] diff 3.5762786865234375e-06
ay[0], by[0] allclose False
这个问题在Linux + V100 + torch1.9.0 + cu111上进行了测试,但到目前为止,许多其他配置也出现了这个问题。有什么线索吗?或者只是我误解了conv2d应该如何工作?
当我使用批次大小为1验证我的trainset结果时,我遇到了这个问题,但它与我记录的训练过程中的错误明显不同,所以我进行了检查,发现是conv2d层引起了这个问题。如果我正确理解conv2d,这不应该发生。
英文:
I feed the conv2d with the same data but different batch size (using stack) as input:
a = torch.rand(1, 512, 16, 16) # (1, 512, 16, 16)
b = torch.cat([a, a, a], dim=0) # (3, 512, 16, 16)
a, b = a.cuda(), b.cuda()
net = nn.Conv2d(512, 1024, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
net = net.cuda()
ay = net(a)
by = net(b)
print('ay[0], by[0] max diff', torch.max(torch.abs(ay[0] - by[0])).item())
print('ay[0], by[0] allclose', torch.allclose(ay[0], by[0]))
the result is however different:
ay[0], by[0] diff 3.5762786865234375e-06
ay[0], by[0] allclose False
this problem is tested on Linux + V100 + torch1.9.0 + cu111, but so far many other configuration also seen such problem. Any clue why? Or it is just simply I misunderstand how conv2d should work?
I run into this problem when I validate my trainset result using batch size 1, but it is significantly different from the error I recorded from training process, so I checked for it and find that it is the conv2d layer that causes this problem. If I understand conv2d correctly, this should not be happening.
答案1
得分: 2
据我所知,这个问题与con2d操作无关,而是由于浮点数精度有限引起的,这可以根据操作和架构而变化。这是一个已知的问题,例如,请参阅pytorch-forum上的讨论。
您当前运行的GPU计算可能正在使用单精度浮点计算,如果将其设置为双精度,则误差差异应该会减小:
torch.set_default_tensor_type(torch.DoubleTensor)
这意味着:
print('ay[0], by[0] allclose', torch.allclose(ay[0], by[0], atol=1e-6))
应该打印:
ay[0], by[0] allclose True
至少在我的情况下,在Linux上使用A100进行测试时是这样的。
英文:
As far as I know the problem is not specific to con2d operations, but rather do to a limited floating point precision which can vary depending on the operations and architecture. This is a known issue, see e.g. this discussion on the pytorch-forum.
The GPU calculations you are currently running is probably using single-precision float computations, if you set it to be double-precision the error discrepancies should be reduced:
torch.set_default_tensor_type(torch.DoubleTensor)
meaning that:
print('ay[0], by[0] allclose', torch.allclose(ay[0], by[0], atol=1e-6))
Should print:
> ay[0], by[0] allclose True
At least this is the case for me when testing also on Linux using an A100.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库,让每个人都能够通过互相帮助和分享经验来进步。
评论