refine and align with pytorch

[hjchen2]

• conv1d/2d/3d和deconv1d/2d/3d kernel，stride和dilation参数支持 int 传参
• 提供Tensor.zero_()接口
• Tensor.norm 参数命名
• flow.max和flow.min用法
• flow.nn.functional.dropout，支持inplace

[Flowingsun007]

max看torch文档好像还有一个接口（输入输出都是一个tensor的那种？）：
torch.max(input) → [Tensor](https://pytorch.org/docs/stable/tensors.html#torch.Tensor)

[hjchen2]

"TensorTuple (Tensor input, Int32 dim=None, Bool keepdim=False) => Max", 这个就支持你说的那种，输出如果是TensorTuple，而且size是1的话，就自动会取第0个Tensor作为输出

[Flowingsun007]

哦哦好的，忽略

[MARD1NO]

我们还有个地方是flow.zero_但是接口是flow.zeros_，也顺便改下吧

上下文在：#7573

[hjchen2]

done，见commit 204dd3f

[github-actions]

Static analysis with clang failed. PR label automerge has been removed

[github-actions]

CI failed when running job: cpu-module. PR label automerge has been removed

[github-actions]

Speed stats:

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.7ms (= 12871.1ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 140.9ms (= 14090.3ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.09 (= 140.9ms / 128.7ms)

✔️ OneFlow resnet50 time: 78.3ms (= 7832.9ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.0ms (= 8502.5ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.09 (= 85.0ms / 78.3ms)

OneFlow resnet50 time: 53.5ms (= 10705.1ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 56.0ms (= 11196.2ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.05 (= 56.0ms / 53.5ms)

OneFlow resnet50 time: 44.0ms (= 8795.9ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 50.2ms (= 10030.4ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.14 (= 50.2ms / 44.0ms)

OneFlow resnet50 time: 37.4ms (= 7481.1ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 44.4ms (= 8886.0ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.19 (= 44.4ms / 37.4ms)

✔️ OneFlow resnet50 time: 142.1ms (= 14210.0ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 160.8ms (= 16075.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.13 (= 160.8ms / 142.1ms)

OneFlow resnet50 time: 90.9ms (= 9094.7ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 105.9ms (= 10588.9ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.16 (= 105.9ms / 90.9ms)

OneFlow resnet50 time: 62.9ms (= 12583.0ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 77.5ms (= 15500.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.23 (= 77.5ms / 62.9ms)

OneFlow resnet50 time: 52.1ms (= 10425.1ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.8ms (= 13566.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.30 (= 67.8ms / 52.1ms)

OneFlow resnet50 time: 48.5ms (= 9691.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 62.8ms (= 12565.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.30 (= 62.8ms / 48.5ms)