Fea/nn graph/block proxy func

[strint]

nn.Graph支持nn.Module的forward中调用了自定义函数的情况

[strint]

主要改动在这里，其它的是代码的整理、简化

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

PyTorch resnet50 time: 137.7ms (= 6887.0ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 125.8ms (= 6289.8ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
Relative speed: 1.09 (= 137.7ms / 125.8ms)

PyTorch resnet50 time: 85.0ms (= 4250.4ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 72.7ms (= 3635.8ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
Relative speed: 1.17 (= 85.0ms / 72.7ms)

PyTorch resnet50 time: 56.5ms (= 2822.6ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 47.1ms (= 2353.2ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
Relative speed: 1.20 (= 56.5ms / 47.1ms)

PyTorch resnet50 time: 49.0ms (= 2449.6ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 41.7ms (= 2083.9ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
Relative speed: 1.18 (= 49.0ms / 41.7ms)

PyTorch resnet50 time: 38.7ms (= 1935.0ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 44.4ms (= 2221.2ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
Relative speed: 0.87 (= 38.7ms / 44.4ms)