Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix permute wrong test case #8083

Merged
merged 3 commits into from
Apr 24, 2022
Merged

Fix permute wrong test case #8083

merged 3 commits into from
Apr 24, 2022

Conversation

MARD1NO
Copy link
Contributor

@MARD1NO MARD1NO commented Apr 24, 2022

No description provided.

@MARD1NO MARD1NO marked this pull request as ready for review April 24, 2022 02:38
@github-actions
Copy link
Contributor

Speed stats:
GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.8ms (= 12880.1ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 144.5ms (= 14452.4ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.12 (= 144.5ms / 128.8ms)

OneFlow resnet50 time: 80.4ms (= 8038.9ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 82.4ms (= 8240.4ms / 100, input_shape=[8, 3, 224, 224])
❌ Relative speed: 1.03 (= 82.4ms / 80.4ms)

OneFlow resnet50 time: 54.8ms (= 10961.2ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 58.2ms (= 11647.8ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.06 (= 58.2ms / 54.8ms)

OneFlow resnet50 time: 46.2ms (= 9245.1ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 41.7ms (= 8337.5ms / 200, input_shape=[2, 3, 224, 224])
❌ Relative speed: 0.90 (= 41.7ms / 46.2ms)

OneFlow resnet50 time: 37.9ms (= 7573.0ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 36.1ms (= 7222.3ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 0.95 (= 36.1ms / 37.9ms)

OneFlow swin dataloader time: 0.264s (= 52.783s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.885s / 200, num_workers=1)
Relative speed: 0.566 (= 0.149s / 0.264s)

OneFlow swin dataloader time: 0.068s (= 13.646s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.162s / 200, num_workers=4)
Relative speed: 0.598 (= 0.041s / 0.068s)

OneFlow swin dataloader time: 0.038s (= 7.552s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.446s / 200, num_workers=8)
Relative speed: 0.589 (= 0.022s / 0.038s)

✔️ OneFlow resnet50 time: 136.2ms (= 13618.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 167.4ms (= 16735.6ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.23 (= 167.4ms / 136.2ms)

OneFlow resnet50 time: 91.8ms (= 9176.6ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.5ms (= 10151.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.11 (= 101.5ms / 91.8ms)

OneFlow resnet50 time: 62.8ms (= 12565.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 83.5ms (= 16698.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.33 (= 83.5ms / 62.8ms)

OneFlow resnet50 time: 55.0ms (= 11009.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 71.0ms (= 14196.9ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.29 (= 71.0ms / 55.0ms)

OneFlow resnet50 time: 50.6ms (= 10118.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 73.7ms (= 14738.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.46 (= 73.7ms / 50.6ms)

@oneflow-ci-bot oneflow-ci-bot requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 24, 2022 06:05
@github-actions
Copy link
Contributor

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8083/

@github-actions
Copy link
Contributor

Speed stats:
GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.1ms (= 12909.7ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.3ms (= 14233.2ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 142.3ms / 129.1ms)

OneFlow resnet50 time: 81.4ms (= 8140.1ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.4ms (= 8439.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.04 (= 84.4ms / 81.4ms)

OneFlow resnet50 time: 54.1ms (= 10815.3ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 60.7ms (= 12136.1ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.12 (= 60.7ms / 54.1ms)

OneFlow resnet50 time: 44.1ms (= 8829.1ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 41.4ms (= 8272.8ms / 200, input_shape=[2, 3, 224, 224])
❌ Relative speed: 0.94 (= 41.4ms / 44.1ms)

OneFlow resnet50 time: 40.9ms (= 8183.9ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 36.0ms (= 7200.1ms / 200, input_shape=[1, 3, 224, 224])
❌ Relative speed: 0.88 (= 36.0ms / 40.9ms)

OneFlow swin dataloader time: 0.265s (= 52.924s / 200, num_workers=1)
PyTorch swin dataloader time: 0.151s (= 30.111s / 200, num_workers=1)
Relative speed: 0.569 (= 0.151s / 0.265s)

OneFlow swin dataloader time: 0.068s (= 13.607s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.255s / 200, num_workers=4)
Relative speed: 0.607 (= 0.041s / 0.068s)

OneFlow swin dataloader time: 0.038s (= 7.621s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.371s / 200, num_workers=8)
Relative speed: 0.574 (= 0.022s / 0.038s)

✔️ OneFlow resnet50 time: 136.1ms (= 13605.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.3ms (= 16133.1ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.19 (= 161.3ms / 136.1ms)

OneFlow resnet50 time: 90.1ms (= 9014.7ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.3ms (= 10129.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.12 (= 101.3ms / 90.1ms)

OneFlow resnet50 time: 62.3ms (= 12457.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 78.0ms (= 15603.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.25 (= 78.0ms / 62.3ms)

OneFlow resnet50 time: 54.5ms (= 10901.0ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.8ms (= 13965.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.28 (= 69.8ms / 54.5ms)

OneFlow resnet50 time: 52.0ms (= 10404.6ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 73.5ms (= 14695.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.41 (= 73.5ms / 52.0ms)

@oneflow-ci-bot oneflow-ci-bot requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 24, 2022 06:22
@github-actions
Copy link
Contributor

Speed stats:
GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.2ms (= 12918.5ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.4ms (= 14238.3ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 142.4ms / 129.2ms)

OneFlow resnet50 time: 80.8ms (= 8084.4ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.4ms (= 8442.8ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.04 (= 84.4ms / 80.8ms)

OneFlow resnet50 time: 54.8ms (= 10964.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 55.9ms (= 11180.8ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.02 (= 55.9ms / 54.8ms)

OneFlow resnet50 time: 44.7ms (= 8945.6ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 40.7ms (= 8130.2ms / 200, input_shape=[2, 3, 224, 224])
❌ Relative speed: 0.91 (= 40.7ms / 44.7ms)

OneFlow resnet50 time: 40.3ms (= 8066.4ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 35.9ms (= 7175.4ms / 200, input_shape=[1, 3, 224, 224])
❌ Relative speed: 0.89 (= 35.9ms / 40.3ms)

OneFlow swin dataloader time: 0.256s (= 51.114s / 200, num_workers=1)
PyTorch swin dataloader time: 0.152s (= 30.343s / 200, num_workers=1)
Relative speed: 0.594 (= 0.152s / 0.256s)

OneFlow swin dataloader time: 0.069s (= 13.804s / 200, num_workers=4)
PyTorch swin dataloader time: 0.044s (= 8.745s / 200, num_workers=4)
Relative speed: 0.633 (= 0.044s / 0.069s)

OneFlow swin dataloader time: 0.038s (= 7.552s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.396s / 200, num_workers=8)
Relative speed: 0.582 (= 0.022s / 0.038s)

✔️ OneFlow resnet50 time: 136.1ms (= 13609.4ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.8ms (= 16175.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.19 (= 161.8ms / 136.1ms)

OneFlow resnet50 time: 91.3ms (= 9129.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.7ms (= 10171.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.11 (= 101.7ms / 91.3ms)

OneFlow resnet50 time: 63.3ms (= 12650.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 77.3ms (= 15462.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.22 (= 77.3ms / 63.3ms)

OneFlow resnet50 time: 53.2ms (= 10638.1ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.1ms (= 13618.3ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.28 (= 68.1ms / 53.2ms)

OneFlow resnet50 time: 51.9ms (= 10377.6ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 75.3ms (= 15052.7ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.45 (= 75.3ms / 51.9ms)

@github-actions
Copy link
Contributor

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8083/

@mergify mergify bot merged commit b76ad20 into master Apr 24, 2022
@mergify mergify bot deleted the fix_permute_test branch April 24, 2022 07:48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants