add is_cpu api

[hjchen2]

No description provided.

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 141.4ms (= 14135.0ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 147.4ms (= 14736.9ms / 100, input_shape=[16, 3, 224, 224])
❌ Relative speed: 1.04 (= 147.4ms / 141.4ms)

OneFlow resnet50 time: 83.0ms (= 8296.5ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 94.7ms (= 9471.1ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.14 (= 94.7ms / 83.0ms)

OneFlow resnet50 time: 51.8ms (= 10353.5ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 73.8ms (= 14765.3ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.43 (= 73.8ms / 51.8ms)

OneFlow resnet50 time: 34.2ms (= 6837.6ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 66.1ms (= 13218.7ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.93 (= 66.1ms / 34.2ms)

OneFlow resnet50 time: 26.3ms (= 5256.3ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 57.8ms (= 11565.8ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 2.20 (= 57.8ms / 26.3ms)

OneFlow swin dataloader time: 0.241s (= 48.175s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.702s / 200, num_workers=1)
Relative speed: 0.617 (= 0.149s / 0.241s)

OneFlow swin dataloader time: 0.067s (= 13.481s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.216s / 200, num_workers=4)
Relative speed: 0.609 (= 0.041s / 0.067s)

OneFlow swin dataloader time: 0.043s (= 8.638s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.455s / 200, num_workers=8)
Relative speed: 0.516 (= 0.022s / 0.043s)

❌ OneFlow resnet50 time: 154.4ms (= 15443.1ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 164.2ms (= 16422.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
❌ Relative speed: 1.06 (= 164.2ms / 154.4ms)

OneFlow resnet50 time: 94.2ms (= 9420.4ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 104.6ms (= 10462.5ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.11 (= 104.6ms / 94.2ms)

OneFlow resnet50 time: 62.1ms (= 12422.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 79.8ms (= 15965.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.29 (= 79.8ms / 62.1ms)

OneFlow resnet50 time: 44.3ms (= 8860.0ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 71.4ms (= 14279.1ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.61 (= 71.4ms / 44.3ms)

OneFlow resnet50 time: 36.9ms (= 7387.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 66.9ms (= 13389.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.81 (= 66.9ms / 36.9ms)

[github-actions]

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/10172/