disable eager_op consistent mode

[lixinqi]

禁止动态shape op的consistent模式

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

PyTorch resnet50 time: 139.8ms (= 6991.8ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 126.1ms (= 6303.1ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
Relative speed: 1.11 (= 139.8ms / 126.1ms)

PyTorch resnet50 time: 83.3ms (= 4165.8ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 75.9ms (= 3793.6ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
Relative speed: 1.10 (= 83.3ms / 75.9ms)

PyTorch resnet50 time: 58.7ms (= 2935.7ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 48.5ms (= 2423.2ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
Relative speed: 1.21 (= 58.7ms / 48.5ms)

PyTorch resnet50 time: 48.3ms (= 2412.6ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 41.2ms (= 2059.7ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
Relative speed: 1.17 (= 48.3ms / 41.2ms)

PyTorch resnet50 time: 44.3ms (= 2215.1ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 41.7ms (= 2086.9ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
Relative speed: 1.06 (= 44.3ms / 41.7ms)