-
Notifications
You must be signed in to change notification settings - Fork 6.8k
[v1.5.x] [MKLDNN] Independent gradients requests check with respect to weights… #15805
Conversation
f2060c0
to
8686c44
Compare
@zixuanweeei This PR is cherrypicking your commit merged on master branch for v1.5.1, could you please review? Thanks. |
@@ -1906,6 +1906,88 @@ def test_depthwise_convolution(): | |||
for arr1, arr2 in zip(exe1.outputs + exe1.grad_arrays, exe2.outputs + exe2.grad_arrays): | |||
np.testing.assert_allclose(arr1.asnumpy(), arr2.asnumpy(), rtol=1e-3, atol=1e-3) | |||
|
|||
|
|||
@with_seed() | |||
def test_convolution_independent_gradients(): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We have reported flakiness on this UT and added some notes. Seems the the cherry-picked commit is behind the merged one on master branch. And it looks good on master. Would you mind checking it again? Thanks. https://github.com/apache/incubator-mxnet/blob/57927a932610581a351e4aed4fb7c9209d0e57c2/tests/python/unittest/test_operator.py#L2003-L2006
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry for a mistake. The flakiness is fixed by #15631. I think these two should be taken into v1.5.x together.
aafe905
to
1420bc0
Compare
@zixuanweeei thanks for the information, the one for test case changes also taking, please kindly review. |
… and bias of convolution (apache#15497) * Independent req[kBias] and req[kWeight] check * Add UT for independent conv gradient requests * Update conv independent grad UT with no_bias enabled * Check req[kWeight] for avoiding unnecessary prim registration * Check `OpReqTpye` in CommitOutput automatically * Lock cudnn autotune for accurate conv output * Ignore independent gradients test on GPU * Trigger CI * Sets a low bar for autotuned cudnn convolution
…ients (apache#15631) * Skip test_convolution_independent_gradirents * Add an issue link * Fix inconsistent context of input array and binding op * Trigger CI * Retrigger CI
LGTM. Thanks. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
… and bias of convolution (#15497)
Independent req[kBias] and req[kWeight] check
Add UT for independent conv gradient requests
Update conv independent grad UT with no_bias enabled
Check req[kWeight] for avoiding unnecessary prim registration
Check
OpReqTpye
in CommitOutput automaticallyLock cudnn autotune for accurate conv output
Ignore independent gradients test on GPU
Trigger CI
Sets a low bar for autotuned cudnn convolution
Description
cherry-picking PR#15497
Checklist
Essentials
Please feel free to remove inapplicable items for your PR.
Changes
Comments