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

concat type promotion #7921

Merged
merged 32 commits into from
Apr 4, 2022
Merged

concat type promotion #7921

merged 32 commits into from
Apr 4, 2022

Conversation

Flowingsun007
Copy link
Contributor

@Flowingsun007 Flowingsun007 commented Mar 29, 2022

@github-actions
Copy link
Contributor

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

@Flowingsun007 Flowingsun007 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 3, 2022 11:10
@Flowingsun007 Flowingsun007 removed the request for review from oneflow-ci-bot April 3, 2022 17:13
@oneflow-ci-bot oneflow-ci-bot requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 3, 2022 19:32
@github-actions
Copy link
Contributor

github-actions bot commented Apr 3, 2022

CI failed when running job: cuda-module-distributed-rank-1. PR label automerge has been removed

@github-actions github-actions bot removed the automerge label Apr 3, 2022
@github-actions
Copy link
Contributor

github-actions bot commented Apr 3, 2022

CI failed when running job: cuda-module-distributed-rank-0. PR label automerge has been removed

@Flowingsun007 Flowingsun007 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 4, 2022 02:21
@github-actions
Copy link
Contributor

github-actions bot commented Apr 4, 2022

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/7921/

@Flowingsun007 Flowingsun007 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 4, 2022 07:56
@Flowingsun007 Flowingsun007 added the need-highest-priority Only add this when you really need it!!! Will block all other PRs. label Apr 4, 2022
@github-actions
Copy link
Contributor

github-actions bot commented Apr 4, 2022

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/7921/

@github-actions
Copy link
Contributor

github-actions bot commented Apr 4, 2022

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

@github-actions github-actions bot removed the automerge label Apr 4, 2022
@Flowingsun007 Flowingsun007 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot April 4, 2022 13:53
@github-actions
Copy link
Contributor

github-actions bot commented Apr 4, 2022

Speed stats:
GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.1ms (= 12810.4ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 137.7ms (= 13772.0ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.08 (= 137.7ms / 128.1ms)

✔️ OneFlow resnet50 time: 78.7ms (= 7871.5ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 83.8ms (= 8379.1ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.06 (= 83.8ms / 78.7ms)

OneFlow resnet50 time: 50.4ms (= 10085.0ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 55.3ms (= 11056.7ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.10 (= 55.3ms / 50.4ms)

OneFlow resnet50 time: 42.7ms (= 8530.5ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 48.2ms (= 9644.6ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.13 (= 48.2ms / 42.7ms)

OneFlow resnet50 time: 39.0ms (= 7791.3ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 38.4ms (= 7672.2ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 0.98 (= 38.4ms / 39.0ms)

OneFlow swin dataloader time: 0.251s (= 50.160s / 200, num_workers=1)
PyTorch swin dataloader time: 0.249s (= 49.770s / 200, num_workers=1)
✔️ Relative speed: 0.992 (= 0.249s / 0.251s)

OneFlow swin dataloader time: 0.064s (= 12.827s / 200, num_workers=4)
PyTorch swin dataloader time: 0.067s (= 13.405s / 200, num_workers=4)
✔️ Relative speed: 1.045 (= 0.067s / 0.064s)

OneFlow swin dataloader time: 0.038s (= 7.580s / 200, num_workers=8)
PyTorch swin dataloader time: 0.039s (= 7.753s / 200, num_workers=8)
✔️ Relative speed: 1.023 (= 0.039s / 0.038s)

✔️ OneFlow resnet50 time: 135.5ms (= 13552.7ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 155.8ms (= 15584.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.15 (= 155.8ms / 135.5ms)

OneFlow resnet50 time: 87.6ms (= 8763.6ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 98.9ms (= 9886.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.13 (= 98.9ms / 87.6ms)

OneFlow resnet50 time: 58.5ms (= 11701.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 75.0ms (= 14991.5ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.28 (= 75.0ms / 58.5ms)

OneFlow resnet50 time: 53.2ms (= 10632.2ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 66.6ms (= 13310.2ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.25 (= 66.6ms / 53.2ms)

OneFlow resnet50 time: 50.6ms (= 10118.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 62.2ms (= 12440.7ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.23 (= 62.2ms / 50.6ms)

@github-actions
Copy link
Contributor

github-actions bot commented Apr 4, 2022

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/7921/

@Flowingsun007 Flowingsun007 merged commit 14efa94 into master Apr 4, 2022
@Flowingsun007 Flowingsun007 deleted the concat_type_promotion branch April 4, 2022 14:31
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug eager enhancement need-highest-priority Only add this when you really need it!!! Will block all other PRs.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants