CI test time(not merge)

[lixiang007666]

[github-actions[bot]]

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.7ms (= 12867.8ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.0ms (= 14204.2ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 142.0ms / 128.7ms)

OneFlow resnet50 time: 75.8ms (= 7580.1ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 86.0ms (= 8596.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.13 (= 86.0ms / 75.8ms)

OneFlow resnet50 time: 49.5ms (= 9901.0ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 60.2ms (= 12030.3ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.22 (= 60.2ms / 49.5ms)

OneFlow resnet50 time: 36.8ms (= 7351.5ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 42.0ms (= 8399.8ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.14 (= 42.0ms / 36.8ms)

OneFlow resnet50 time: 31.0ms (= 6205.1ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 45.5ms (= 9096.6ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.47 (= 45.5ms / 31.0ms)

OneFlow swin dataloader time: 0.259s (= 51.867s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 30.001s / 200, num_workers=1)
Relative speed: 0.578 (= 0.150s / 0.259s)

OneFlow swin dataloader time: 0.070s (= 14.061s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.203s / 200, num_workers=4)
Relative speed: 0.583 (= 0.041s / 0.070s)

OneFlow swin dataloader time: 0.040s (= 7.957s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.442s / 200, num_workers=8)
Relative speed: 0.558 (= 0.022s / 0.040s)

❌ OneFlow resnet50 time: 137.2ms (= 13721.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 164.8ms (= 16479.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 164.8ms / 137.2ms)

OneFlow resnet50 time: 86.3ms (= 8629.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 112.6ms (= 11262.4ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.31 (= 112.6ms / 86.3ms)

OneFlow resnet50 time: 58.1ms (= 11615.0ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 89.1ms (= 17817.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.53 (= 89.1ms / 58.1ms)

OneFlow resnet50 time: 44.5ms (= 8908.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.2ms (= 13649.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.53 (= 68.2ms / 44.5ms)

OneFlow resnet50 time: 38.2ms (= 7635.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.7ms (= 13745.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.80 (= 68.7ms / 38.2ms)

[github-actions[bot]]

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8820/

[github-actions[bot]]

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.3ms (= 12831.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.6ms (= 14260.5ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.11 (= 142.6ms / 128.3ms)

OneFlow resnet50 time: 75.2ms (= 7522.2ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.2ms (= 8518.2ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.13 (= 85.2ms / 75.2ms)

OneFlow resnet50 time: 48.4ms (= 9688.4ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 61.7ms (= 12344.1ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.27 (= 61.7ms / 48.4ms)

OneFlow resnet50 time: 36.1ms (= 7213.1ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 41.6ms (= 8321.3ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.15 (= 41.6ms / 36.1ms)

OneFlow resnet50 time: 28.2ms (= 5634.3ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 39.0ms (= 7801.9ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.38 (= 39.0ms / 28.2ms)

OneFlow swin dataloader time: 0.269s (= 53.897s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.887s / 200, num_workers=1)
Relative speed: 0.555 (= 0.149s / 0.269s)

OneFlow swin dataloader time: 0.112s (= 22.376s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.200s / 200, num_workers=4)
Relative speed: 0.366 (= 0.041s / 0.112s)

OneFlow swin dataloader time: 0.041s (= 8.197s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.399s / 200, num_workers=8)
Relative speed: 0.537 (= 0.022s / 0.041s)

❌ OneFlow resnet50 time: 136.7ms (= 13665.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 159.0ms (= 15896.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.16 (= 159.0ms / 136.7ms)

OneFlow resnet50 time: 84.1ms (= 8409.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.8ms (= 10179.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.21 (= 101.8ms / 84.1ms)

OneFlow resnet50 time: 57.3ms (= 11464.7ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 79.7ms (= 15943.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.39 (= 79.7ms / 57.3ms)

OneFlow resnet50 time: 45.3ms (= 9051.2ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 72.6ms (= 14523.2ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.60 (= 72.6ms / 45.3ms)

OneFlow resnet50 time: 39.0ms (= 7795.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 76.5ms (= 15300.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.96 (= 76.5ms / 39.0ms)