nn.Graph with training

[strint]

• train nn.Linear with loss
• deal with MultiClientContext delete after sync vm
• deal with eager inplace
• deal with 0 dim tensor
• deal with op default attr
• train Alexnet

[github-actions]

CI failed, removing label automerge

[github-actions]

CI failed, removing label automerge

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

PyTorch resnet50 time: 142.1ms (= 7105.0ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 126.7ms (= 6337.1ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
Relative speed: 1.12 (= 142.1ms / 126.7ms)

PyTorch resnet50 time: 84.8ms (= 4242.1ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 74.1ms (= 3707.0ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
Relative speed: 1.14 (= 84.8ms / 74.1ms)

PyTorch resnet50 time: 59.4ms (= 2970.9ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 47.5ms (= 2376.9ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
Relative speed: 1.25 (= 59.4ms / 47.5ms)

PyTorch resnet50 time: 51.5ms (= 2576.6ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 48.6ms (= 2427.8ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
Relative speed: 1.06 (= 51.5ms / 48.6ms)

PyTorch resnet50 time: 44.9ms (= 2243.4ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 42.3ms (= 2112.7ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
Relative speed: 1.06 (= 44.9ms / 42.3ms)