Remove IDMgr::GetGpuPhyIdFromThrdId/IDMgr::GetDeviceTypeFromThrdId

[liujuncheng]

No description provided.

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

OneFlow resnet50 time: 128.7ms (= 6433.5ms / 50, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 144.0ms (= 7201.0ms / 50, input_shape=[16, 3, 224, 224])
Relative speed: 1.12 (= 144.0ms / 128.7ms)

OneFlow resnet50 time: 74.8ms (= 3738.2ms / 50, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 86.3ms (= 4314.9ms / 50, input_shape=[8, 3, 224, 224])
Relative speed: 1.15 (= 86.3ms / 74.8ms)

OneFlow resnet50 time: 48.9ms (= 2444.3ms / 50, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 58.6ms (= 2931.9ms / 50, input_shape=[4, 3, 224, 224])
Relative speed: 1.20 (= 58.6ms / 48.9ms)

OneFlow resnet50 time: 41.0ms (= 2050.7ms / 50, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 49.8ms (= 2489.6ms / 50, input_shape=[2, 3, 224, 224])
Relative speed: 1.21 (= 49.8ms / 41.0ms)

OneFlow resnet50 time: 38.9ms (= 1944.2ms / 50, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 41.7ms (= 2082.7ms / 50, input_shape=[1, 3, 224, 224])
Relative speed: 1.07 (= 41.7ms / 38.9ms)

OneFlow resnet50 time: 154.1ms (= 7705.4ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 160.4ms (= 8019.3ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
Relative speed: 1.04 (= 160.4ms / 154.1ms)

OneFlow resnet50 time: 101.7ms (= 5082.7ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 99.7ms (= 4984.6ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
Relative speed: 0.98 (= 99.7ms / 101.7ms)

OneFlow resnet50 time: 83.8ms (= 4188.3ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 85.4ms (= 4268.7ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
Relative speed: 1.02 (= 85.4ms / 83.8ms)

OneFlow resnet50 time: 69.3ms (= 3466.8ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 62.3ms (= 3116.0ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
Relative speed: 0.90 (= 62.3ms / 69.3ms)

OneFlow resnet50 time: 69.7ms (= 3486.0ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 59.9ms (= 2993.2ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
Relative speed: 0.86 (= 59.9ms / 69.7ms)

[leaves-zwx]

这里为什么不是直接成员？因为 TaskId 不能随着 TaskNode 的初始化而初始化，且 TaskId 无默认构造函数吗？

[liujuncheng]

这里为什么不是直接成员？因为 TaskId 不能随着 TaskNode 的初始化而初始化，且 TaskId 无默认构造函数吗？

是

[github-actions]

CI failed, removing label automerge

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

OneFlow resnet50 time: 128.0ms (= 6400.7ms / 50, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 140.9ms (= 7046.1ms / 50, input_shape=[16, 3, 224, 224])
Relative speed: 1.10 (= 140.9ms / 128.0ms)

OneFlow resnet50 time: 74.5ms (= 3725.1ms / 50, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 86.9ms (= 4346.4ms / 50, input_shape=[8, 3, 224, 224])
Relative speed: 1.17 (= 86.9ms / 74.5ms)

OneFlow resnet50 time: 47.7ms (= 2383.4ms / 50, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 62.8ms (= 3139.9ms / 50, input_shape=[4, 3, 224, 224])
Relative speed: 1.32 (= 62.8ms / 47.7ms)

OneFlow resnet50 time: 44.3ms (= 2215.5ms / 50, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 45.3ms (= 2267.5ms / 50, input_shape=[2, 3, 224, 224])
Relative speed: 1.02 (= 45.3ms / 44.3ms)

OneFlow resnet50 time: 40.4ms (= 2018.7ms / 50, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 45.3ms (= 2266.4ms / 50, input_shape=[1, 3, 224, 224])
Relative speed: 1.12 (= 45.3ms / 40.4ms)

OneFlow resnet50 time: 158.4ms (= 7917.9ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 160.5ms (= 8027.4ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
Relative speed: 1.01 (= 160.5ms / 158.4ms)

OneFlow resnet50 time: 101.1ms (= 5053.4ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 100.1ms (= 5003.7ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
Relative speed: 0.99 (= 100.1ms / 101.1ms)

OneFlow resnet50 time: 74.0ms (= 3702.0ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 74.0ms (= 3700.6ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
Relative speed: 1.00 (= 74.0ms / 74.0ms)

OneFlow resnet50 time: 73.6ms (= 3677.5ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.3ms (= 3364.9ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
Relative speed: 0.91 (= 67.3ms / 73.6ms)

OneFlow resnet50 time: 68.7ms (= 3432.9ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 60.6ms (= 3029.8ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
Relative speed: 0.88 (= 60.6ms / 68.7ms)