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.

Sign up for GitHub

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

Jump to bottom

Fix stack bug for 129inputs #8927

Merged
merged 7 commits into from
Aug 17, 2022
Merged

Fix stack bug for 129inputs #8927

merged 7 commits into from
Aug 17, 2022

Conversation

wyg1997
Copy link
Contributor

@wyg1997 wyg1997 commented Aug 16, 2022

fix #8918

@wyg1997 wyg1997 requested review from MARD1NO and Flowingsun007 and removed request for jackalcooper, daquexian, hjchen2 and BBuf August 16, 2022 09:10
@wyg1997 wyg1997 requested a review from oneflow-ci-bot August 16, 2022 09:27
@github-actions
Copy link
Contributor

Static analysis with clang failed. PR label automerge has been removed

@wyg1997 wyg1997 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot August 16, 2022 10:46
@github-actions
Copy link
Contributor

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8927/

@github-actions
Copy link
Contributor

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

@github-actions
Copy link
Contributor

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8927/

@wyg1997 wyg1997 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot August 17, 2022 06:41
@github-actions
Copy link
Contributor

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8927/

@github-actions
Copy link
Contributor

Speed stats: 
GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.3ms (= 12830.4ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 148.6ms (= 14856.6ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.16 (= 148.6ms / 128.3ms)

OneFlow resnet50 time: 75.5ms (= 7552.0ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.0ms (= 8397.5ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.11 (= 84.0ms / 75.5ms)

OneFlow resnet50 time: 48.5ms (= 9698.2ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 55.5ms (= 11093.2ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.14 (= 55.5ms / 48.5ms)

OneFlow resnet50 time: 36.0ms (= 7199.3ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 46.0ms (= 9195.7ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.28 (= 46.0ms / 36.0ms)

OneFlow resnet50 time: 28.1ms (= 5627.0ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 40.0ms (= 7995.9ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.42 (= 40.0ms / 28.1ms)

OneFlow swin dataloader time: 0.270s (= 53.913s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 30.063s / 200, num_workers=1)
Relative speed: 0.558 (= 0.150s / 0.270s)

OneFlow swin dataloader time: 0.072s (= 14.397s / 200, num_workers=4)
PyTorch swin dataloader time: 0.043s (= 8.648s / 200, num_workers=4)
Relative speed: 0.601 (= 0.043s / 0.072s)

OneFlow swin dataloader time: 0.041s (= 8.258s / 200, num_workers=8)
PyTorch swin dataloader time: 0.021s (= 4.242s / 200, num_workers=8)
Relative speed: 0.514 (= 0.021s / 0.041s)

❌ OneFlow resnet50 time: 136.7ms (= 13672.8ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 162.3ms (= 16232.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.19 (= 162.3ms / 136.7ms)

OneFlow resnet50 time: 84.8ms (= 8482.6ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 102.9ms (= 10285.9ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.21 (= 102.9ms / 84.8ms)

OneFlow resnet50 time: 58.3ms (= 11654.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 83.9ms (= 16783.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.44 (= 83.9ms / 58.3ms)

OneFlow resnet50 time: 45.2ms (= 9030.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.6ms (= 13725.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.52 (= 68.6ms / 45.2ms)

OneFlow resnet50 time: 38.8ms (= 7768.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 68.0ms (= 13608.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.75 (= 68.0ms / 38.8ms)

@mergify mergify bot merged commit 4a33514 into master Aug 17, 2022
@mergify mergify bot deleted the fix-stack_bug_for_129inputs branch August 17, 2022 09:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@MARD1NO MARD1NO MARD1NO approved these changes

@Flowingsun007 Flowingsun007 Flowingsun007 approved these changes

@oneflow-ci-bot oneflow-ci-bot Awaiting requested review from oneflow-ci-bot

Assignees
No one assigned
Labels
automerge bug eager
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

[BUG] 128*n+1 个 Tensor 使用 flow.stack 时后向出错
3 participants
@wyg1997 @Flowingsun007 @MARD1NO