Add check meta args in to global api

[clackhan]

根据 https://github.com/Oneflow-Inc/OneTeam/issues/1311#issuecomment-1098711629 讨论结果，添加tensor.local_to_global/tensor.global_to_global接口，并给tensor.to_global接口添加check_meta参数。

• tensor.local_to_global

• tensor.global_to_global

• tensor.to_global

[github-actions]

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8027/

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.4ms (= 12837.9ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 149.9ms (= 14987.2ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.17 (= 149.9ms / 128.4ms)

OneFlow resnet50 time: 80.2ms (= 8018.0ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.9ms (= 8588.7ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.07 (= 85.9ms / 80.2ms)

OneFlow resnet50 time: 53.9ms (= 10781.5ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 58.4ms (= 11684.6ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.08 (= 58.4ms / 53.9ms)

OneFlow resnet50 time: 45.1ms (= 9016.7ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 44.4ms (= 8883.6ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 0.99 (= 44.4ms / 45.1ms)

OneFlow resnet50 time: 41.0ms (= 8200.9ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 35.4ms (= 7075.7ms / 200, input_shape=[1, 3, 224, 224])
❌ Relative speed: 0.86 (= 35.4ms / 41.0ms)

OneFlow swin dataloader time: 0.258s (= 51.569s / 200, num_workers=1)
PyTorch swin dataloader time: 0.153s (= 30.661s / 200, num_workers=1)
Relative speed: 0.595 (= 0.153s / 0.258s)

OneFlow swin dataloader time: 0.069s (= 13.759s / 200, num_workers=4)
PyTorch swin dataloader time: 0.042s (= 8.367s / 200, num_workers=4)
Relative speed: 0.608 (= 0.042s / 0.069s)

OneFlow swin dataloader time: 0.038s (= 7.683s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.473s / 200, num_workers=8)
Relative speed: 0.582 (= 0.022s / 0.038s)

❌ OneFlow resnet50 time: 140.4ms (= 14044.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 160.9ms (= 16086.0ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.15 (= 160.9ms / 140.4ms)

OneFlow resnet50 time: 86.0ms (= 8601.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.2ms (= 10119.0ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.18 (= 101.2ms / 86.0ms)

OneFlow resnet50 time: 65.3ms (= 13050.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 80.8ms (= 16161.7ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.24 (= 80.8ms / 65.3ms)

OneFlow resnet50 time: 54.6ms (= 10927.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.9ms (= 13974.9ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.28 (= 69.9ms / 54.6ms)

OneFlow resnet50 time: 49.6ms (= 9923.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.6ms (= 13514.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.36 (= 67.6ms / 49.6ms)

[github-actions]

CI failed when running job: cuda-module. PR label automerge has been removed

[github-actions]

Speed stats:

[github-actions]

Speed stats:

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.5ms (= 12852.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 150.5ms (= 15046.8ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.17 (= 150.5ms / 128.5ms)

OneFlow resnet50 time: 80.6ms (= 8059.1ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 86.0ms (= 8601.1ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.07 (= 86.0ms / 80.6ms)

OneFlow resnet50 time: 53.8ms (= 10768.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 61.4ms (= 12285.5ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.14 (= 61.4ms / 53.8ms)

OneFlow resnet50 time: 43.4ms (= 8682.2ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 41.0ms (= 8196.1ms / 200, input_shape=[2, 3, 224, 224])
❌ Relative speed: 0.94 (= 41.0ms / 43.4ms)

OneFlow resnet50 time: 42.3ms (= 8450.2ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 33.8ms (= 6754.6ms / 200, input_shape=[1, 3, 224, 224])
❌ Relative speed: 0.80 (= 33.8ms / 42.3ms)

OneFlow swin dataloader time: 0.257s (= 51.403s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 29.916s / 200, num_workers=1)
Relative speed: 0.582 (= 0.150s / 0.257s)

OneFlow swin dataloader time: 0.066s (= 13.167s / 200, num_workers=4)
PyTorch swin dataloader time: 0.042s (= 8.365s / 200, num_workers=4)
Relative speed: 0.635 (= 0.042s / 0.066s)

OneFlow swin dataloader time: 0.037s (= 7.476s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.463s / 200, num_workers=8)
Relative speed: 0.597 (= 0.022s / 0.037s)

✔️ OneFlow resnet50 time: 136.0ms (= 13599.8ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 162.5ms (= 16253.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 162.5ms / 136.0ms)

OneFlow resnet50 time: 89.6ms (= 8957.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.3ms (= 10133.0ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.13 (= 101.3ms / 89.6ms)

OneFlow resnet50 time: 63.1ms (= 12625.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 79.2ms (= 15841.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.25 (= 79.2ms / 63.1ms)

OneFlow resnet50 time: 55.6ms (= 11119.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 80.8ms (= 16166.0ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.45 (= 80.8ms / 55.6ms)

OneFlow resnet50 time: 50.9ms (= 10176.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.2ms (= 13636.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.34 (= 68.2ms / 50.9ms)

[github-actions]

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8027/