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

re: transfer oneflow_compile from onediff to oneflow #10408

Merged
merged 20 commits into from
Jan 18, 2024

Conversation

linzs148
Copy link
Contributor

resubmit for PR #10404

Copy link
Contributor

mergify bot commented Jan 18, 2024

⚠️ The sha of the head commit of this PR conflicts with #10404. Mergify cannot evaluate rules on this PR. ⚠️

Copy link
Contributor

Copy link
Contributor

CI failed when running job: cpu-misc. PR label automerge has been removed

Copy link
Contributor

Copy link
Contributor

Copy link
Contributor

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

Copy link
Contributor

Copy link
Contributor

Copy link
Contributor

Speed stats:
GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.7ms (= 4374.7ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 57.5ms (= 5747.4ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.31 (= 57.5ms / 43.7ms)

OneFlow resnet50 time: 26.1ms (= 2610.4ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 37.8ms (= 3778.0ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.45 (= 37.8ms / 26.1ms)

OneFlow resnet50 time: 17.7ms (= 3538.6ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 34.7ms (= 6944.8ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.96 (= 34.7ms / 17.7ms)

OneFlow resnet50 time: 16.8ms (= 3351.0ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 34.7ms (= 6943.3ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 2.07 (= 34.7ms / 16.8ms)

OneFlow resnet50 time: 17.3ms (= 3450.5ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 28.8ms (= 5768.9ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.67 (= 28.8ms / 17.3ms)

OneFlow swin dataloader time: 0.201s (= 40.141s / 200, num_workers=1)
PyTorch swin dataloader time: 0.127s (= 25.342s / 200, num_workers=1)
Relative speed: 0.631 (= 0.127s / 0.201s)

OneFlow swin dataloader time: 0.057s (= 11.492s / 200, num_workers=4)
PyTorch swin dataloader time: 0.032s (= 6.422s / 200, num_workers=4)
Relative speed: 0.559 (= 0.032s / 0.057s)

OneFlow swin dataloader time: 0.030s (= 6.076s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.316s / 200, num_workers=8)
Relative speed: 0.546 (= 0.017s / 0.030s)

❌ OneFlow resnet50 time: 48.8ms (= 4883.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 65.7ms (= 6565.4ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.34 (= 65.7ms / 48.8ms)

OneFlow resnet50 time: 36.9ms (= 3694.3ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 46.1ms (= 4607.9ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.25 (= 46.1ms / 36.9ms)

OneFlow resnet50 time: 28.2ms (= 5630.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 40.0ms (= 8001.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.42 (= 40.0ms / 28.2ms)

OneFlow resnet50 time: 25.1ms (= 5016.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 38.8ms (= 7752.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.55 (= 38.8ms / 25.1ms)

OneFlow resnet50 time: 24.1ms (= 4825.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 36.5ms (= 7298.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.51 (= 36.5ms / 24.1ms)

@linzs148 linzs148 merged commit 2e331e1 into master Jan 18, 2024
20 checks passed
@linzs148 linzs148 deleted the feat/transfer_oneflow_compile branch January 18, 2024 16:32
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