Fix make of_format fail in ubuntu

[tingkuanpei]

Fix python open utf-8 file fail problem in Ubuntu when runing make of_format

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

OneFlow resnet50 time: 128.3ms (= 6416.9ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 141.4ms (= 7070.8ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.10 (= 141.4ms / 128.3ms)

OneFlow resnet50 time: 74.8ms (= 3741.0ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 83.2ms (= 4161.4ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.11 (= 83.2ms / 74.8ms)

OneFlow resnet50 time: 48.0ms (= 2399.7ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 58.0ms (= 2898.3ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.21 (= 58.0ms / 48.0ms)

OneFlow resnet50 time: 38.5ms (= 1924.0ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 49.5ms (= 2474.8ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.29 (= 49.5ms / 38.5ms)

OneFlow resnet50 time: 41.6ms (= 2079.8ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 44.2ms (= 2208.7ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.06 (= 44.2ms / 41.6ms)

OneFlow resnet50 time: 139.4ms (= 6971.1ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 150.1ms (= 7506.1ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.08 (= 150.1ms / 139.4ms)

OneFlow resnet50 time: 89.6ms (= 4479.3ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 94.1ms (= 4707.1ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.05 (= 94.1ms / 89.6ms)

OneFlow resnet50 time: 59.0ms (= 2952.5ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 68.0ms (= 3401.7ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 1.15 (= 68.0ms / 59.0ms)

OneFlow resnet50 time: 56.8ms (= 2840.6ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 54.7ms (= 2733.3ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 0.96 (= 54.7ms / 56.8ms)

OneFlow resnet50 time: 49.8ms (= 2492.4ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow GPU used (rank 0): 0 MiB
PyTorch resnet50 time: 45.2ms (= 2261.9ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
PyTorch GPU used (rank 0, estimated): 0 MiB
Relative speed: 0.91 (= 45.2ms / 49.8ms)