support autograd function in graph #8843

marigoold · 2022-08-04T03:52:13Z

autograd.Function自定义op支持graph模式

…rpreter

github-actions · 2022-08-22T04:06:11Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8843/

github-actions · 2022-08-22T04:08:40Z

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.9ms (= 12885.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 143.2ms (= 14316.2ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.11 (= 143.2ms / 128.9ms)

OneFlow resnet50 time: 75.5ms (= 7548.0ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 83.7ms (= 8367.5ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.11 (= 83.7ms / 75.5ms)

OneFlow resnet50 time: 48.6ms (= 9710.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 56.1ms (= 11215.2ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.15 (= 56.1ms / 48.6ms)

OneFlow resnet50 time: 36.0ms (= 7209.5ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 40.8ms (= 8159.3ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.13 (= 40.8ms / 36.0ms)

OneFlow resnet50 time: 28.1ms (= 5622.2ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 35.5ms (= 7104.4ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.26 (= 35.5ms / 28.1ms)

OneFlow swin dataloader time: 0.270s (= 54.014s / 200, num_workers=1)
PyTorch swin dataloader time: 0.154s (= 30.764s / 200, num_workers=1)
Relative speed: 0.570 (= 0.154s / 0.270s)

OneFlow swin dataloader time: 0.073s (= 14.680s / 200, num_workers=4)
PyTorch swin dataloader time: 0.042s (= 8.415s / 200, num_workers=4)
Relative speed: 0.573 (= 0.042s / 0.073s)

OneFlow swin dataloader time: 0.062s (= 12.492s / 200, num_workers=8)
PyTorch swin dataloader time: 0.021s (= 4.278s / 200, num_workers=8)
Relative speed: 0.342 (= 0.021s / 0.062s)

❌ OneFlow resnet50 time: 136.7ms (= 13669.3ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.2ms (= 16119.9ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.18 (= 161.2ms / 136.7ms)

OneFlow resnet50 time: 84.9ms (= 8494.4ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 102.4ms (= 10237.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.21 (= 102.4ms / 84.9ms)

OneFlow resnet50 time: 57.7ms (= 11546.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 88.0ms (= 17590.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.52 (= 88.0ms / 57.7ms)

OneFlow resnet50 time: 45.1ms (= 9011.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 70.1ms (= 14016.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.56 (= 70.1ms / 45.1ms)

OneFlow resnet50 time: 38.9ms (= 7778.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.0ms (= 13790.1ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.77 (= 69.0ms / 38.9ms)

github-actions · 2022-08-22T04:09:43Z

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

github-actions · 2022-08-23T04:33:05Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8843/

github-actions · 2022-08-23T04:34:15Z

Speed stats:

GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.5ms (= 12954.7ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.2ms (= 14222.4ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 142.2ms / 129.5ms)

OneFlow resnet50 time: 76.3ms (= 7628.7ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 89.1ms (= 8905.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.17 (= 89.1ms / 76.3ms)

OneFlow resnet50 time: 48.7ms (= 9738.4ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 54.2ms (= 10845.2ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.11 (= 54.2ms / 48.7ms)

OneFlow resnet50 time: 35.6ms (= 7122.8ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 43.0ms (= 8607.9ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.21 (= 43.0ms / 35.6ms)

OneFlow resnet50 time: 29.6ms (= 5916.2ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 42.4ms (= 8472.8ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.43 (= 42.4ms / 29.6ms)

OneFlow swin dataloader time: 0.257s (= 51.332s / 200, num_workers=1)
PyTorch swin dataloader time: 0.152s (= 30.403s / 200, num_workers=1)
Relative speed: 0.592 (= 0.152s / 0.257s)

OneFlow swin dataloader time: 0.075s (= 15.030s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.189s / 200, num_workers=4)
Relative speed: 0.545 (= 0.041s / 0.075s)

OneFlow swin dataloader time: 0.062s (= 12.464s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.314s / 200, num_workers=8)
Relative speed: 0.346 (= 0.022s / 0.062s)

❌ OneFlow resnet50 time: 139.1ms (= 13908.7ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 164.0ms (= 16403.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.18 (= 164.0ms / 139.1ms)

OneFlow resnet50 time: 86.5ms (= 8652.0ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 103.6ms (= 10355.3ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 103.6ms / 86.5ms)

OneFlow resnet50 time: 58.7ms (= 11744.0ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 80.0ms (= 16005.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.36 (= 80.0ms / 58.7ms)

OneFlow resnet50 time: 45.5ms (= 9099.8ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 71.6ms (= 14329.0ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.57 (= 71.6ms / 45.5ms)

OneFlow resnet50 time: 39.3ms (= 7869.0ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.3ms (= 13459.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.71 (= 67.3ms / 39.3ms)

github-actions · 2022-08-23T04:56:31Z

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

…dev_support_autograd_function_in_graph

github-actions · 2022-08-30T12:35:02Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8843/

github-actions · 2022-08-30T12:35:12Z

Speed stats:

GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.2ms (= 12922.9ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 143.1ms (= 14311.4ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.11 (= 143.1ms / 129.2ms)

OneFlow resnet50 time: 74.3ms (= 7434.1ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.3ms (= 8430.5ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.13 (= 84.3ms / 74.3ms)

OneFlow resnet50 time: 46.8ms (= 9363.2ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 59.0ms (= 11800.5ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.26 (= 59.0ms / 46.8ms)

OneFlow resnet50 time: 34.3ms (= 6869.7ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 42.3ms (= 8463.5ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.23 (= 42.3ms / 34.3ms)

OneFlow resnet50 time: 28.2ms (= 5631.1ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 42.8ms (= 8563.6ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.52 (= 42.8ms / 28.2ms)

OneFlow swin dataloader time: 0.271s (= 54.206s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 29.949s / 200, num_workers=1)
Relative speed: 0.553 (= 0.150s / 0.271s)

OneFlow swin dataloader time: 0.071s (= 14.187s / 200, num_workers=4)
PyTorch swin dataloader time: 0.040s (= 8.080s / 200, num_workers=4)
Relative speed: 0.570 (= 0.040s / 0.071s)

OneFlow swin dataloader time: 0.064s (= 12.842s / 200, num_workers=8)
PyTorch swin dataloader time: 0.023s (= 4.507s / 200, num_workers=8)
Relative speed: 0.351 (= 0.023s / 0.064s)

❌ OneFlow resnet50 time: 138.0ms (= 13802.0ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 162.8ms (= 16283.1ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.18 (= 162.8ms / 138.0ms)

OneFlow resnet50 time: 84.2ms (= 8424.5ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 102.5ms (= 10251.4ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.22 (= 102.5ms / 84.2ms)

OneFlow resnet50 time: 57.3ms (= 11451.2ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 78.6ms (= 15726.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.37 (= 78.6ms / 57.3ms)

OneFlow resnet50 time: 44.1ms (= 8829.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 72.0ms (= 14394.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.63 (= 72.0ms / 44.1ms)

OneFlow resnet50 time: 38.6ms (= 7725.8ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 70.0ms (= 13996.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.81 (= 70.0ms / 38.6ms)

github-actions · 2022-08-30T14:12:44Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8843/

github-actions · 2022-08-30T14:15:22Z

Speed stats:

GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.2ms (= 12920.8ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 150.0ms (= 15003.7ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.16 (= 150.0ms / 129.2ms)

OneFlow resnet50 time: 74.5ms (= 7449.3ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.3ms (= 8430.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.13 (= 84.3ms / 74.5ms)

OneFlow resnet50 time: 46.9ms (= 9387.1ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 58.6ms (= 11725.1ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.25 (= 58.6ms / 46.9ms)

OneFlow resnet50 time: 34.2ms (= 6842.2ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 47.5ms (= 9503.6ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.39 (= 47.5ms / 34.2ms)

OneFlow resnet50 time: 28.2ms (= 5640.0ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 37.9ms (= 7584.4ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.34 (= 37.9ms / 28.2ms)

OneFlow swin dataloader time: 0.259s (= 51.735s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.834s / 200, num_workers=1)
Relative speed: 0.577 (= 0.149s / 0.259s)

OneFlow swin dataloader time: 0.070s (= 13.933s / 200, num_workers=4)
PyTorch swin dataloader time: 0.044s (= 8.780s / 200, num_workers=4)
Relative speed: 0.630 (= 0.044s / 0.070s)

OneFlow swin dataloader time: 0.040s (= 7.973s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.393s / 200, num_workers=8)
Relative speed: 0.551 (= 0.022s / 0.040s)

❌ OneFlow resnet50 time: 138.3ms (= 13831.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.3ms (= 16129.4ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 161.3ms / 138.3ms)

OneFlow resnet50 time: 84.7ms (= 8474.6ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 112.5ms (= 11249.5ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.33 (= 112.5ms / 84.7ms)

OneFlow resnet50 time: 57.7ms (= 11533.6ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 79.2ms (= 15837.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.37 (= 79.2ms / 57.7ms)

OneFlow resnet50 time: 44.6ms (= 8923.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 72.0ms (= 14399.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.61 (= 72.0ms / 44.6ms)

OneFlow resnet50 time: 38.6ms (= 7723.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 78.2ms (= 15641.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 2.03 (= 78.2ms / 38.6ms)

github-actions · 2022-08-30T14:15:49Z

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

github-actions · 2022-08-30T14:39:43Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8843/

github-actions · 2022-08-30T14:42:46Z

Speed stats:

GPU Name: GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.3ms (= 12926.5ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 148.7ms (= 14870.2ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.15 (= 148.7ms / 129.3ms)

OneFlow resnet50 time: 74.6ms (= 7463.2ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.3ms (= 8528.5ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.14 (= 85.3ms / 74.6ms)

OneFlow resnet50 time: 47.5ms (= 9503.7ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 59.5ms (= 11894.8ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.25 (= 59.5ms / 47.5ms)

OneFlow resnet50 time: 34.8ms (= 6955.0ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 43.5ms (= 8696.6ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.25 (= 43.5ms / 34.8ms)

OneFlow resnet50 time: 28.4ms (= 5671.6ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 36.6ms (= 7319.2ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.29 (= 36.6ms / 28.4ms)

OneFlow swin dataloader time: 0.260s (= 52.076s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.803s / 200, num_workers=1)
Relative speed: 0.572 (= 0.149s / 0.260s)

OneFlow swin dataloader time: 0.069s (= 13.861s / 200, num_workers=4)
PyTorch swin dataloader time: 0.040s (= 7.994s / 200, num_workers=4)
Relative speed: 0.577 (= 0.040s / 0.069s)

OneFlow swin dataloader time: 0.040s (= 8.048s / 200, num_workers=8)
PyTorch swin dataloader time: 0.023s (= 4.565s / 200, num_workers=8)
Relative speed: 0.567 (= 0.023s / 0.040s)

❌ OneFlow resnet50 time: 138.2ms (= 13817.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 162.3ms (= 16232.9ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 162.3ms / 138.2ms)

OneFlow resnet50 time: 84.5ms (= 8450.7ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 102.1ms (= 10212.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.21 (= 102.1ms / 84.5ms)

OneFlow resnet50 time: 57.3ms (= 11467.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 78.7ms (= 15739.6ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.37 (= 78.7ms / 57.3ms)

OneFlow resnet50 time: 44.2ms (= 8835.9ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.2ms (= 13840.7ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.57 (= 69.2ms / 44.2ms)

OneFlow resnet50 time: 38.5ms (= 7703.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.8ms (= 13553.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.76 (= 67.8ms / 38.5ms)

marigoold and others added 9 commits August 2, 2022 09:40

add unittest for graph autograd function

5eff8c6

add autograd function support for lazy

2eacc67

add using OpExprInterpreter::ApplyImpl in derived lazy and eager inte…

30f7247

…rpreter

refine code

8cb5c2b

remove applyimpl in base class, implement in lazy/eager interpreter

7bdfb33

refine code

1a6b00c

Merge branch 'master' into dev_support_autograd_function_in_graph

2983401

refine graph autograd function test

8dfc67e

Merge branch 'master' into dev_support_autograd_function_in_graph

7d8d50a

hjchen2 marked this pull request as ready for review August 22, 2022 03:07

hjchen2 requested review from chengtbf, strint, BBuf, daquexian and jackalcooper as code owners August 22, 2022 03:07

hjchen2 requested review from wyg1997 and oneflow-ci-bot August 22, 2022 03:07

hjchen2 added feature automerge graph graph mode labels Aug 22, 2022

hjchen2 enabled auto-merge (squash) August 22, 2022 03:08

hjchen2 approved these changes Aug 22, 2022

View reviewed changes

wyg1997 approved these changes Aug 22, 2022

View reviewed changes

mergify bot and others added 2 commits August 22, 2022 03:10

Merge branch 'master' into dev_support_autograd_function_in_graph

f03e6e4

check weight grad

6c8181a

hjchen2 changed the title ~~[WIP] Dev support autograd function in graph~~ support autograd function in graph Aug 22, 2022

refine

938a2c9

github-actions bot removed the automerge label Aug 22, 2022

Merge branch 'master' into dev_support_autograd_function_in_graph

ec17050

hjchen2 added the automerge label Aug 23, 2022

Merge branch 'master' into dev_support_autograd_function_in_graph

92032b6

github-actions bot removed the automerge label Aug 23, 2022

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

55571ce

…dev_support_autograd_function_in_graph

hjchen2 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot August 30, 2022 12:58

hjchen2 added the automerge label Aug 30, 2022

Merge branch 'master' into dev_support_autograd_function_in_graph

7e0d7d2

github-actions bot removed the automerge label Aug 30, 2022

hjchen2 requested review from oneflow-ci-bot and removed request for oneflow-ci-bot August 30, 2022 14:22

hjchen2 added the automerge label Aug 30, 2022

oneflow-ci-bot requested review from oneflow-ci-bot and removed request for oneflow-ci-bot August 30, 2022 16:04

hjchen2 merged commit adb6376 into master Aug 30, 2022

hjchen2 deleted the dev_support_autograd_function_in_graph branch August 30, 2022 16:41

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

support autograd function in graph #8843

support autograd function in graph #8843

marigoold commented Aug 4, 2022 •

edited by hjchen2

Loading

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

support autograd function in graph #8843

support autograd function in graph #8843

Conversation

marigoold commented Aug 4, 2022 • edited by hjchen2 Loading

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 22, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 23, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

github-actions bot commented Aug 30, 2022

marigoold commented Aug 4, 2022 •

edited by hjchen2

Loading