fix upsample shape infer bug #8105

BBuf · 2022-04-27T02:42:12Z

fix upsample shape infer bug caused by float precision. linked: Oneflow-Inc/libai#260 (comment)

MARD1NO

Add Attr("output_size")

simonJJJ

为什么设置eps=1e-5呢

BBuf · 2022-04-27T05:43:25Z

为什么设置eps=1e-5呢

准备基于output_size重构一下op，先忽略这个改动，会去掉。

…/oneflow into fix_upsample_shape_infer_bug

* fix_upsample_shape_infer_bug * fix 5 nearest * add 5 nearest test * fix 5 nearest test * fix 1 linear * fix 4 bilinear * fix 4 bicubic * modify bicubic 2d file name * fix 5 trilinear * fix exception info * fix exception info * fix bug * modify interpolate * change float to double * rm useless SI64ArrayAttr: in OneFlowUserOps * rm useless import in cpp * update * add judge for output_size * update oneflow/oneflow/core/autograd/gradient_funcs/upsample.cpp * add grad in td * test failed * fix small failed case in upsample * solve test error * change float to double * align to fix_upsample_shape_infer_bug * align to fix_upsample_shape_infer_bug Co-authored-by: BBuf <1182563586@qq.com>

simonJJJ

interpolate.py 103行这个地方还有必要保留scale吗

oneflow/core/functional/impl/array_functor.cpp

oneflow/ir/include/OneFlow/OneFlowUserOps.td

oneflow/user/kernels/upsample_bicubic_2d_kernel.cu

oneflow/user/ops/upsample_op.cpp

oneflow/user/kernels/upsample_bicubic_2d_kernel.cpp

BBuf · 2022-05-11T07:03:15Z

interpolate.py 103行这个地方还有必要保留scale吗

这个暂时去不掉，因为scale不是Optional的，改起来工作量依然比较大，准备下个pr来做。

github-actions · 2022-05-20T17:01:30Z

Speed stats:

github-actions · 2022-05-20T18:54:41Z

Speed stats:

GPU Name: NVIDIA GeForce GTX 1080 

❌ OneFlow resnet50 time: 129.9ms (= 12988.5ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 141.4ms (= 14141.0ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.09 (= 141.4ms / 129.9ms)

OneFlow resnet50 time: 76.6ms (= 7659.6ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 82.8ms (= 8278.9ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.08 (= 82.8ms / 76.6ms)

OneFlow resnet50 time: 49.4ms (= 9887.9ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 55.1ms (= 11024.6ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.11 (= 55.1ms / 49.4ms)

OneFlow resnet50 time: 42.1ms (= 8416.8ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 45.3ms (= 9065.2ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.08 (= 45.3ms / 42.1ms)

OneFlow resnet50 time: 37.8ms (= 7561.5ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 38.9ms (= 7781.6ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.03 (= 38.9ms / 37.8ms)

OneFlow swin dataloader time: 0.247s (= 49.459s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 29.969s / 200, num_workers=1)
Relative speed: 0.606 (= 0.150s / 0.247s)

OneFlow swin dataloader time: 0.066s (= 13.149s / 200, num_workers=4)
PyTorch swin dataloader time: 0.043s (= 8.568s / 200, num_workers=4)
Relative speed: 0.652 (= 0.043s / 0.066s)

OneFlow swin dataloader time: 0.038s (= 7.631s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.454s / 200, num_workers=8)
Relative speed: 0.584 (= 0.022s / 0.038s)

❌ OneFlow resnet50 time: 145.7ms (= 14566.9ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 170.8ms (= 17077.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 170.8ms / 145.7ms)

OneFlow resnet50 time: 97.2ms (= 9720.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 112.2ms (= 11224.3ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.15 (= 112.2ms / 97.2ms)

OneFlow resnet50 time: 74.3ms (= 14858.0ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 87.1ms (= 17414.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
❌ Relative speed: 1.17 (= 87.1ms / 74.3ms)

OneFlow resnet50 time: 61.2ms (= 12235.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 73.6ms (= 14723.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 73.6ms / 61.2ms)

OneFlow resnet50 time: 55.7ms (= 11135.2ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.5ms (= 13892.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.25 (= 69.5ms / 55.7ms)

github-actions · 2022-05-20T19:29:40Z

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

github-actions · 2022-05-21T02:12:22Z

View latest API docs preview at: https://staging.oneflow.info/docs/Oneflow-Inc/oneflow/pr/8105/

github-actions · 2022-05-21T02:19:21Z

Speed stats:

GPU Name: NVIDIA GeForce GTX 1080 

❌ OneFlow resnet50 time: 130.3ms (= 13026.4ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 142.8ms (= 14284.5ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 142.8ms / 130.3ms)

OneFlow resnet50 time: 78.0ms (= 7796.2ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.2ms (= 8419.6ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.08 (= 84.2ms / 78.0ms)

OneFlow resnet50 time: 54.1ms (= 10821.6ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 54.8ms (= 10951.3ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.01 (= 54.8ms / 54.1ms)

OneFlow resnet50 time: 42.3ms (= 8459.3ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 44.1ms (= 8824.2ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.04 (= 44.1ms / 42.3ms)

OneFlow resnet50 time: 37.5ms (= 7503.6ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 39.8ms (= 7967.3ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.06 (= 39.8ms / 37.5ms)

OneFlow swin dataloader time: 0.247s (= 49.358s / 200, num_workers=1)
PyTorch swin dataloader time: 0.152s (= 30.405s / 200, num_workers=1)
Relative speed: 0.616 (= 0.152s / 0.247s)

OneFlow swin dataloader time: 0.065s (= 12.983s / 200, num_workers=4)
PyTorch swin dataloader time: 0.043s (= 8.516s / 200, num_workers=4)
Relative speed: 0.656 (= 0.043s / 0.065s)

OneFlow swin dataloader time: 0.035s (= 6.981s / 200, num_workers=8)
PyTorch swin dataloader time: 0.023s (= 4.565s / 200, num_workers=8)
Relative speed: 0.654 (= 0.023s / 0.035s)

❌ OneFlow resnet50 time: 146.1ms (= 14609.6ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 169.9ms (= 16987.1ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.16 (= 169.9ms / 146.1ms)

OneFlow resnet50 time: 98.2ms (= 9817.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 112.1ms (= 11213.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.14 (= 112.1ms / 98.2ms)

OneFlow resnet50 time: 73.3ms (= 14661.3ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 87.2ms (= 17432.4ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.19 (= 87.2ms / 73.3ms)

OneFlow resnet50 time: 62.2ms (= 12449.5ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 74.6ms (= 14924.3ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 74.6ms / 62.2ms)

OneFlow resnet50 time: 54.2ms (= 10837.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 74.3ms (= 14852.0ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.37 (= 74.3ms / 54.2ms)

github-actions · 2022-05-21T02:53:59Z

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

github-actions · 2022-05-21T04:18:24Z

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.9ms (= 12891.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 140.7ms (= 14065.8ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.09 (= 140.7ms / 128.9ms)

OneFlow resnet50 time: 75.5ms (= 7548.5ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.0ms (= 8504.2ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.13 (= 85.0ms / 75.5ms)

OneFlow resnet50 time: 52.5ms (= 10499.1ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 52.7ms (= 10546.5ms / 200, input_shape=[4, 3, 224, 224])
❌ Relative speed: 1.00 (= 52.7ms / 52.5ms)

OneFlow resnet50 time: 40.4ms (= 8070.2ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 44.3ms (= 8864.3ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.10 (= 44.3ms / 40.4ms)

OneFlow resnet50 time: 37.1ms (= 7428.2ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 38.0ms (= 7601.3ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.02 (= 38.0ms / 37.1ms)

OneFlow swin dataloader time: 0.396s (= 79.116s / 200, num_workers=1)
PyTorch swin dataloader time: 0.150s (= 29.973s / 200, num_workers=1)
Relative speed: 0.379 (= 0.150s / 0.396s)

OneFlow swin dataloader time: 0.062s (= 12.489s / 200, num_workers=4)
PyTorch swin dataloader time: 0.040s (= 7.954s / 200, num_workers=4)
Relative speed: 0.637 (= 0.040s / 0.062s)

OneFlow swin dataloader time: 0.036s (= 7.231s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.397s / 200, num_workers=8)
Relative speed: 0.608 (= 0.022s / 0.036s)

✔️ OneFlow resnet50 time: 136.3ms (= 13625.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 159.8ms (= 15981.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 159.8ms / 136.3ms)

OneFlow resnet50 time: 88.4ms (= 8835.9ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 101.1ms (= 10113.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.14 (= 101.1ms / 88.4ms)

OneFlow resnet50 time: 60.6ms (= 12110.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 79.0ms (= 15805.8ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.31 (= 79.0ms / 60.6ms)

OneFlow resnet50 time: 51.1ms (= 10215.3ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 69.1ms (= 13812.2ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.35 (= 69.1ms / 51.1ms)

OneFlow resnet50 time: 45.1ms (= 9010.0ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 67.2ms (= 13430.4ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.49 (= 67.2ms / 45.1ms)

github-actions · 2022-05-21T05:13:09Z

Speed stats:

GPU Name: GeForce GTX 1080 

✔️ OneFlow resnet50 time: 128.9ms (= 12888.3ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 141.6ms (= 14157.6ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.10 (= 141.6ms / 128.9ms)

OneFlow resnet50 time: 75.9ms (= 7591.4ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 82.2ms (= 8216.4ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.08 (= 82.2ms / 75.9ms)

OneFlow resnet50 time: 51.5ms (= 10292.8ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 63.9ms (= 12790.0ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.24 (= 63.9ms / 51.5ms)

OneFlow resnet50 time: 41.5ms (= 8291.0ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 40.6ms (= 8110.4ms / 200, input_shape=[2, 3, 224, 224])
❌ Relative speed: 0.98 (= 40.6ms / 41.5ms)

OneFlow resnet50 time: 37.0ms (= 7396.3ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 37.5ms (= 7499.4ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.01 (= 37.5ms / 37.0ms)

OneFlow swin dataloader time: 0.243s (= 48.628s / 200, num_workers=1)
PyTorch swin dataloader time: 0.149s (= 29.794s / 200, num_workers=1)
Relative speed: 0.613 (= 0.149s / 0.243s)

OneFlow swin dataloader time: 0.064s (= 12.742s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.286s / 200, num_workers=4)
Relative speed: 0.650 (= 0.041s / 0.064s)

OneFlow swin dataloader time: 0.036s (= 7.188s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.318s / 200, num_workers=8)
Relative speed: 0.601 (= 0.022s / 0.036s)

❌ OneFlow resnet50 time: 137.8ms (= 13781.8ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.8ms (= 16179.1ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 161.8ms / 137.8ms)

OneFlow resnet50 time: 86.9ms (= 8690.5ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 102.1ms (= 10208.2ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 102.1ms / 86.9ms)

OneFlow resnet50 time: 62.5ms (= 12491.9ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 88.3ms (= 17659.1ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.41 (= 88.3ms / 62.5ms)

OneFlow resnet50 time: 51.9ms (= 10389.3ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 70.0ms (= 13999.9ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.35 (= 70.0ms / 51.9ms)

OneFlow resnet50 time: 46.3ms (= 9264.5ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 75.6ms (= 15120.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.63 (= 75.6ms / 46.3ms)

github-actions · 2022-05-21T05:49:46Z

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

github-actions · 2022-05-21T14:56:53Z

Speed stats:

GPU Name: NVIDIA GeForce GTX 1080 

❌ OneFlow resnet50 time: 130.2ms (= 13023.1ms / 100, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 144.0ms (= 14404.0ms / 100, input_shape=[16, 3, 224, 224])
✔️ Relative speed: 1.11 (= 144.0ms / 130.2ms)

OneFlow resnet50 time: 76.4ms (= 7636.7ms / 100, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 84.8ms (= 8478.6ms / 100, input_shape=[8, 3, 224, 224])
✔️ Relative speed: 1.11 (= 84.8ms / 76.4ms)

OneFlow resnet50 time: 54.0ms (= 10805.7ms / 200, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 56.1ms (= 11214.0ms / 200, input_shape=[4, 3, 224, 224])
✔️ Relative speed: 1.04 (= 56.1ms / 54.0ms)

OneFlow resnet50 time: 43.0ms (= 8594.0ms / 200, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 45.6ms (= 9126.0ms / 200, input_shape=[2, 3, 224, 224])
✔️ Relative speed: 1.06 (= 45.6ms / 43.0ms)

OneFlow resnet50 time: 37.6ms (= 7529.0ms / 200, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 40.7ms (= 8132.3ms / 200, input_shape=[1, 3, 224, 224])
✔️ Relative speed: 1.08 (= 40.7ms / 37.6ms)

OneFlow swin dataloader time: 0.379s (= 75.843s / 200, num_workers=1)
PyTorch swin dataloader time: 0.151s (= 30.246s / 200, num_workers=1)
Relative speed: 0.399 (= 0.151s / 0.379s)

OneFlow swin dataloader time: 0.065s (= 12.969s / 200, num_workers=4)
PyTorch swin dataloader time: 0.041s (= 8.118s / 200, num_workers=4)
Relative speed: 0.626 (= 0.041s / 0.065s)

OneFlow swin dataloader time: 0.037s (= 7.306s / 200, num_workers=8)
PyTorch swin dataloader time: 0.022s (= 4.435s / 200, num_workers=8)
Relative speed: 0.607 (= 0.022s / 0.037s)

❌ OneFlow resnet50 time: 145.6ms (= 14559.5ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 170.4ms (= 17042.2ms / 100, input_shape=[16, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.17 (= 170.4ms / 145.6ms)

OneFlow resnet50 time: 98.5ms (= 9850.1ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 112.2ms (= 11221.8ms / 100, input_shape=[8, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.14 (= 112.2ms / 98.5ms)

OneFlow resnet50 time: 73.4ms (= 14682.2ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 88.5ms (= 17708.6ms / 200, input_shape=[4, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.21 (= 88.5ms / 73.4ms)

OneFlow resnet50 time: 61.8ms (= 12362.4ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 74.4ms (= 14882.6ms / 200, input_shape=[2, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.20 (= 74.4ms / 61.8ms)

OneFlow resnet50 time: 57.5ms (= 11506.3ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 77.5ms (= 15497.9ms / 200, input_shape=[1, 3, 224, 224], ddp, world size=2)
✔️ Relative speed: 1.35 (= 77.5ms / 57.5ms)

fix upsample shape infer bug

f8ecec7

BBuf requested review from daquexian and jackalcooper as code owners April 27, 2022 02:42

add more example

1246d1c

MARD1NO requested changes Apr 27, 2022

View reviewed changes

simonJJJ reviewed Apr 27, 2022

View reviewed changes

BBuf and others added 5 commits April 28, 2022 20:41

fix eager free tensor bug when in job

0592b57

merge

eb3710d

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

0f3caad

Merge branch 'master' into fix_upsample_shape_infer_bug

6d4d6ac

fix nearest2d bug

1ac4a29

BBuf requested review from hjchen2, liujuncheng and guo-ran as code owners May 6, 2022 06:57

BBuf added 2 commits May 6, 2022 15:31

fix upsample nearest1d shape infer bug

fe6ae1d

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

347d3f1

zhongshsh mentioned this pull request May 7, 2022

Fix upsample shape infer bug continue #8159

Merged

BBuf and others added 8 commits May 9, 2022 10:31

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

2949cd0

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

2dc2375

Merge branch 'master' into fix_upsample_shape_infer_bug

312183f

restruct upsample op

0221e50

Merge branch 'master' into fix_upsample_shape_infer_bug

a3a3c96

change all float scale to double

06a2cb9

Merge branch 'fix_upsample_shape_infer_bug' of github.com:Oneflow-Inc…

173165a

…/oneflow into fix_upsample_shape_infer_bug

simonJJJ reviewed May 11, 2022

View reviewed changes

MARD1NO reviewed May 11, 2022

View reviewed changes

fix comment

7d6bc71

github-actions bot removed the automerge label May 20, 2022

relax speed test

b1686a8

BBuf added the automerge label May 20, 2022

Merge branch 'master' into fix_upsample_shape_infer_bug

89b0379

Merge branch 'master' into fix_upsample_shape_infer_bug

c73a1ec

github-actions bot removed the automerge label May 20, 2022

BBuf added the automerge label May 21, 2022

Merge branch 'master' into fix_upsample_shape_infer_bug

21424da

BBuf requested review from oneflow-ci-bot and removed request for oneflow-ci-bot May 21, 2022 01:36

github-actions bot removed the automerge label May 21, 2022

simplify test case

b9d78ae

BBuf added the automerge label May 21, 2022

BBuf requested review from oneflow-ci-bot and removed request for oneflow-ci-bot May 21, 2022 04:38

github-actions bot removed the automerge label May 21, 2022

BBuf and others added 2 commits May 21, 2022 22:04

fix bug

e8d37f5

Merge branch 'master' into fix_upsample_shape_infer_bug

2c47e2e

BBuf merged commit 0afc09f into master May 21, 2022

BBuf deleted the fix_upsample_shape_infer_bug branch May 21, 2022 15:25

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix upsample shape infer bug #8105

fix upsample shape infer bug #8105

BBuf commented Apr 27, 2022

MARD1NO left a comment

simonJJJ left a comment

BBuf commented Apr 27, 2022 •

edited

Loading

simonJJJ left a comment •

edited

Loading

BBuf commented May 11, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

fix upsample shape infer bug #8105

fix upsample shape infer bug #8105

Conversation

BBuf commented Apr 27, 2022

MARD1NO left a comment

Choose a reason for hiding this comment

simonJJJ left a comment

Choose a reason for hiding this comment

BBuf commented Apr 27, 2022 • edited Loading

simonJJJ left a comment • edited Loading

Choose a reason for hiding this comment

BBuf commented May 11, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 20, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

github-actions bot commented May 21, 2022

BBuf commented Apr 27, 2022 •

edited

Loading

simonJJJ left a comment •

edited

Loading