Fix dynamo mock error #10318

strint · 2023-08-23T04:50:21Z

No description provided.

…fix_dynamo_mock_error

github-actions · 2023-08-23T04:51:52Z

Code got formatted by CI. Please request CI again if you still want to have this PR merged. If the PR is from a forked repo, please download the patch files from the GitHub Actions web page and apply them locally.

github-actions · 2023-08-23T05:22:50Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/10318/

github-actions · 2023-08-23T05:27:10Z

Speed stats:

GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.8ms (= 4382.1ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 57.3ms (= 5729.8ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.31 (= 57.3ms / 43.8ms)

OneFlow resnet50 time: 25.8ms (= 2582.7ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 37.7ms (= 3767.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.46 (= 37.7ms / 25.8ms)

OneFlow resnet50 time: 19.0ms (= 3804.2ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 36.9ms (= 7372.4ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.94 (= 36.9ms / 19.0ms)

OneFlow resnet50 time: 18.4ms (= 3672.1ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 31.8ms (= 6351.6ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.73 (= 31.8ms / 18.4ms)

OneFlow resnet50 time: 17.8ms (= 3569.4ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 29.4ms (= 5883.8ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.65 (= 29.4ms / 17.8ms)

OneFlow swin dataloader time: 0.202s (= 40.441s / 200, num_workers=1)
PyTorch swin dataloader time: 0.128s (= 25.684s / 200, num_workers=1)
Relative speed: 0.635 (= 0.128s / 0.202s)

OneFlow swin dataloader time: 0.055s (= 11.097s / 200, num_workers=4)
PyTorch swin dataloader time: 0.034s (= 6.729s / 200, num_workers=4)
Relative speed: 0.606 (= 0.034s / 0.055s)

OneFlow swin dataloader time: 0.030s (= 5.927s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.323s / 200, num_workers=8)
Relative speed: 0.561 (= 0.017s / 0.030s)

❌ OneFlow resnet50 time: 47.5ms (= 4755.0ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 63.2ms (= 6317.6ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.33 (= 63.2ms / 47.5ms)

OneFlow resnet50 time: 31.8ms (= 3176.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 45.1ms (= 4505.7ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.42 (= 45.1ms / 31.8ms)

OneFlow resnet50 time: 23.9ms (= 4786.2ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 40.8ms (= 8166.5ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.71 (= 40.8ms / 23.9ms)

OneFlow resnet50 time: 21.3ms (= 4254.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 36.7ms (= 7349.1ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.73 (= 36.7ms / 21.3ms)

OneFlow resnet50 time: 21.0ms (= 4205.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 35.7ms (= 7135.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.70 (= 35.7ms / 21.0ms)

…c/oneflow into fix_dynamo_mock_error

github-actions · 2023-08-23T16:42:46Z

Code got formatted by CI. Please request CI again if you still want to have this PR merged. If the PR is from a forked repo, please download the patch files from the GitHub Actions web page and apply them locally.

github-actions · 2023-08-23T17:12:16Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/10318/

github-actions · 2023-08-23T17:17:14Z

Speed stats:

GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.7ms (= 4365.2ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 57.7ms (= 5771.1ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.32 (= 57.7ms / 43.7ms)

OneFlow resnet50 time: 25.9ms (= 2590.4ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 37.6ms (= 3756.8ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.45 (= 37.6ms / 25.9ms)

OneFlow resnet50 time: 19.2ms (= 3838.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 36.3ms (= 7261.3ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.89 (= 36.3ms / 19.2ms)

OneFlow resnet50 time: 18.7ms (= 3747.8ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 34.0ms (= 6808.8ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.82 (= 34.0ms / 18.7ms)

OneFlow resnet50 time: 18.2ms (= 3641.1ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 29.6ms (= 5927.5ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.63 (= 29.6ms / 18.2ms)

OneFlow swin dataloader time: 0.202s (= 40.381s / 200, num_workers=1)
PyTorch swin dataloader time: 0.128s (= 25.653s / 200, num_workers=1)
Relative speed: 0.635 (= 0.128s / 0.202s)

OneFlow swin dataloader time: 0.054s (= 10.845s / 200, num_workers=4)
PyTorch swin dataloader time: 0.033s (= 6.639s / 200, num_workers=4)
Relative speed: 0.612 (= 0.033s / 0.054s)

OneFlow swin dataloader time: 0.031s (= 6.275s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.332s / 200, num_workers=8)
Relative speed: 0.531 (= 0.017s / 0.031s)

❌ OneFlow resnet50 time: 47.6ms (= 4764.8ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 64.8ms (= 6479.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.36 (= 64.8ms / 47.6ms)

OneFlow resnet50 time: 30.8ms (= 3082.0ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 45.7ms (= 4571.6ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.48 (= 45.7ms / 30.8ms)

OneFlow resnet50 time: 24.0ms (= 4800.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 41.8ms (= 8366.6ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.74 (= 41.8ms / 24.0ms)

OneFlow resnet50 time: 22.0ms (= 4405.3ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 37.3ms (= 7452.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.69 (= 37.3ms / 22.0ms)

OneFlow resnet50 time: 21.1ms (= 4212.1ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 34.2ms (= 6837.1ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.62 (= 34.2ms / 21.1ms)

…c/oneflow into fix_dynamo_mock_error

github-actions · 2023-08-24T04:44:45Z

Code got formatted by CI. Please request CI again if you still want to have this PR merged. If the PR is from a forked repo, please download the patch files from the GitHub Actions web page and apply them locally.

github-actions · 2023-08-24T05:16:01Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/10318/

github-actions · 2023-08-24T05:19:56Z

Speed stats:

GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.6ms (= 4360.5ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 61.0ms (= 6097.6ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.40 (= 61.0ms / 43.6ms)

OneFlow resnet50 time: 26.1ms (= 2611.0ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 38.0ms (= 3795.1ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.45 (= 38.0ms / 26.1ms)

OneFlow resnet50 time: 18.8ms (= 3754.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 34.9ms (= 6977.7ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.86 (= 34.9ms / 18.8ms)

OneFlow resnet50 time: 18.4ms (= 3675.2ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 33.1ms (= 6624.1ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.80 (= 33.1ms / 18.4ms)

OneFlow resnet50 time: 17.9ms (= 3584.8ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 28.4ms (= 5676.4ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.58 (= 28.4ms / 17.9ms)

OneFlow swin dataloader time: 0.202s (= 40.363s / 200, num_workers=1)
PyTorch swin dataloader time: 0.128s (= 25.621s / 200, num_workers=1)
Relative speed: 0.635 (= 0.128s / 0.202s)

OneFlow swin dataloader time: 0.053s (= 10.684s / 200, num_workers=4)
PyTorch swin dataloader time: 0.033s (= 6.592s / 200, num_workers=4)
Relative speed: 0.617 (= 0.033s / 0.053s)

OneFlow swin dataloader time: 0.031s (= 6.271s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.379s / 200, num_workers=8)
Relative speed: 0.539 (= 0.017s / 0.031s)

❌ OneFlow resnet50 time: 48.0ms (= 4800.4ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 63.2ms (= 6323.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.32 (= 63.2ms / 48.0ms)

OneFlow resnet50 time: 30.5ms (= 3051.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 44.8ms (= 4476.0ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.47 (= 44.8ms / 30.5ms)

OneFlow resnet50 time: 24.0ms (= 4790.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 41.7ms (= 8336.0ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.74 (= 41.7ms / 24.0ms)

OneFlow resnet50 time: 21.6ms (= 4310.7ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 37.5ms (= 7499.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.74 (= 37.5ms / 21.6ms)

OneFlow resnet50 time: 21.5ms (= 4291.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 34.1ms (= 6815.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.59 (= 34.1ms / 21.5ms)

…c/oneflow into fix_dynamo_mock_error

github-actions · 2023-08-24T10:41:18Z

Speed stats:

github-actions · 2023-08-24T12:24:30Z

Speed stats:

github-actions · 2023-08-28T08:55:00Z

Speed stats:

GPU Name: NVIDIA GeForce RTX 3080 Ti 

❌ OneFlow resnet50 time: 43.7ms (= 4372.9ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 61.1ms (= 6114.4ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.40 (= 61.1ms / 43.7ms)

OneFlow resnet50 time: 25.9ms (= 2587.3ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 37.2ms (= 3716.8ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.44 (= 37.2ms / 25.9ms)

OneFlow resnet50 time: 18.8ms (= 3757.1ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 35.5ms (= 7104.6ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.89 (= 35.5ms / 18.8ms)

OneFlow resnet50 time: 18.1ms (= 3614.1ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 32.2ms (= 6434.0ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.78 (= 32.2ms / 18.1ms)

OneFlow resnet50 time: 18.1ms (= 3619.6ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 30.5ms (= 6098.3ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.68 (= 30.5ms / 18.1ms)

OneFlow swin dataloader time: 0.200s (= 39.996s / 200, num_workers=1)
PyTorch swin dataloader time: 0.129s (= 25.900s / 200, num_workers=1)
Relative speed: 0.648 (= 0.129s / 0.200s)

OneFlow swin dataloader time: 0.055s (= 11.019s / 200, num_workers=4)
PyTorch swin dataloader time: 0.032s (= 6.410s / 200, num_workers=4)
Relative speed: 0.582 (= 0.032s / 0.055s)

OneFlow swin dataloader time: 0.030s (= 6.029s / 200, num_workers=8)
PyTorch swin dataloader time: 0.017s (= 3.301s / 200, num_workers=8)
Relative speed: 0.548 (= 0.017s / 0.030s)

❌ OneFlow resnet50 time: 47.7ms (= 4769.7ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 63.9ms (= 6390.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.34 (= 63.9ms / 47.7ms)

OneFlow resnet50 time: 31.8ms (= 3179.3ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 44.4ms (= 4439.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.40 (= 44.4ms / 31.8ms)

OneFlow resnet50 time: 24.1ms (= 4813.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 40.0ms (= 8008.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.66 (= 40.0ms / 24.1ms)

OneFlow resnet50 time: 21.8ms (= 4351.9ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 36.5ms (= 7301.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.68 (= 36.5ms / 21.8ms)

OneFlow resnet50 time: 20.8ms (= 4163.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 35.5ms (= 7092.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.70 (= 35.5ms / 20.8ms)

github-actions · 2023-08-29T11:03:13Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/10318/

jackalcooper · 2023-08-29T11:23:46Z

python/oneflow/framework/args_tree.py

        elif (
            _is_raw_type(self._value, dict)
            or _is_raw_type(self._value, OrderedDict)
            or _is_raw_type(self._value, list)
            or _is_raw_type(self._value, tuple)
        ):
-            pass
+            repr_str += ", value: " + repr(self._value)


这个和最后的else看着没区别了吗？

这个和最后的else看着没区别了吗？

嗯，处理方式是一样的，先保留了判断逻辑，后面要区分处理可以复用下。

strint and others added 3 commits August 23, 2023 04:47

fake dynamo

30435f9

Merge branch 'master' of https://github.com/Oneflow-Inc/oneflow into …

6cfa893

…fix_dynamo_mock_error

auto format by CI

69bbc0e

strint mentioned this pull request Aug 23, 2023

TypeError: <class 'oneflow.mock_torch.DummyModule'> is not a Module subclass siliconflow/onediff#232

Closed

strint and others added 3 commits August 23, 2023 16:39

general argstree

998c501

Merge branch 'fix_dynamo_mock_error' of https://github.com/Oneflow-In…

eccb5f1

…c/oneflow into fix_dynamo_mock_error

auto format by CI

44a3710

strint and others added 3 commits August 24, 2023 04:42

refine repr of argstree

3619406

Merge branch 'fix_dynamo_mock_error' of https://github.com/Oneflow-In…

93d0fe3

…c/oneflow into fix_dynamo_mock_error

auto format by CI

ccac35e

strint added 2 commits August 24, 2023 10:07

graph allow any input

025d4c5

Merge branch 'fix_dynamo_mock_error' of https://github.com/Oneflow-In…

679e0d5

…c/oneflow into fix_dynamo_mock_error

save support any input

354cf21

strint added graph graph mode bug feature labels Aug 24, 2023

strint mentioned this pull request Aug 24, 2023

sd2 oneflow compile siliconflow/onediff#244

Merged

strint mentioned this pull request Aug 25, 2023

[For dev]Sdxl dev #10323

Closed

Merge branch 'master' into fix_dynamo_mock_error

2788237

hjchen2 approved these changes Aug 28, 2023

View reviewed changes

hjchen2 enabled auto-merge (squash) August 28, 2023 08:14

strint added the automerge label Aug 29, 2023

strint requested a review from oneflow-ci-bot August 29, 2023 10:38

hjchen2 merged commit e4118c7 into master Aug 29, 2023
37 of 39 checks passed

hjchen2 deleted the fix_dynamo_mock_error branch August 29, 2023 11:12

jackalcooper reviewed Aug 29, 2023

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix dynamo mock error #10318

Fix dynamo mock error #10318

strint commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 28, 2023

github-actions bot commented Aug 29, 2023

jackalcooper Aug 29, 2023

strint Aug 29, 2023

Fix dynamo mock error #10318

Fix dynamo mock error #10318

Conversation

strint commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 23, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 24, 2023

github-actions bot commented Aug 28, 2023

github-actions bot commented Aug 29, 2023

jackalcooper Aug 29, 2023

Choose a reason for hiding this comment

strint Aug 29, 2023

Choose a reason for hiding this comment