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

Update dependency constraints on cudnn #10515

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

jackalcooper
Copy link
Collaborator

No description provided.

Copy link
Contributor

github-actions bot commented May 6, 2024

Copy link
Contributor

github-actions bot commented May 6, 2024

Speed stats:
GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.9ms (= 4385.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 57.6ms (= 5762.7ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.31 (= 57.6ms / 43.9ms)

OneFlow resnet50 time: 26.5ms (= 2647.1ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 37.8ms (= 3776.0ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.43 (= 37.8ms / 26.5ms)

OneFlow resnet50 time: 18.5ms (= 3699.5ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 35.1ms (= 7029.0ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.90 (= 35.1ms / 18.5ms)

OneFlow resnet50 time: 17.3ms (= 3451.0ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 31.5ms (= 6292.7ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.82 (= 31.5ms / 17.3ms)

OneFlow resnet50 time: 17.2ms (= 3436.8ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 31.0ms (= 6190.6ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.80 (= 31.0ms / 17.2ms)

OneFlow swin dataloader time: 0.201s (= 40.170s / 200, num_workers=1)
PyTorch swin dataloader time: 0.128s (= 25.600s / 200, num_workers=1)
Relative speed: 0.637 (= 0.128s / 0.201s)

OneFlow swin dataloader time: 0.055s (= 10.995s / 200, num_workers=4)
PyTorch swin dataloader time: 0.033s (= 6.586s / 200, num_workers=4)
Relative speed: 0.599 (= 0.033s / 0.055s)

OneFlow swin dataloader time: 0.031s (= 6.162s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.331s / 200, num_workers=8)
Relative speed: 0.541 (= 0.017s / 0.031s)

❌ OneFlow resnet50 time: 49.3ms (= 4929.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 64.4ms (= 6444.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.31 (= 64.4ms / 49.3ms)

OneFlow resnet50 time: 36.1ms (= 3610.4ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 45.9ms (= 4588.7ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.27 (= 45.9ms / 36.1ms)

OneFlow resnet50 time: 27.5ms (= 5502.7ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 38.5ms (= 7702.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.40 (= 38.5ms / 27.5ms)

OneFlow resnet50 time: 25.1ms (= 5024.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 40.7ms (= 8139.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.62 (= 40.7ms / 25.1ms)

OneFlow resnet50 time: 24.8ms (= 4962.6ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 36.1ms (= 7227.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.46 (= 36.1ms / 24.8ms)

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.

1 participant