[eigen-commits] commit/eigen: 3 new changesets |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen-commits Archives
]
- To: eigen-commits@xxxxxxxxxxxxxxxxxxx
- Subject: [eigen-commits] commit/eigen: 3 new changesets
- From: Bitbucket <commits-noreply@xxxxxxxxxxxxx>
- Date: Wed, 05 Dec 2018 18:19:37 +0000 (UTC)
- Dkim-signature: v=1; a=rsa-sha1; c=relaxed/relaxed; d=bitbucket.org; h=mime-version:content-type:content-transfer-encoding:subject:from:to; s=s1; bh=EgGFQbVDPzjOUf6t4/6a0Y6+EOg=; b=ZedT+top6UDpA/Atr+WAIlC nwoXhaFKgYGf+v5pKOhLMOsi3O3VlSlCsG/+WOS9n/jw6/MNStDWlQRQDUFnUyF4 BDcIq64q8ybDEDfG0LPAdXow9LD4l2Ztj32gi5Id9e7VYEmCI4nL42zApUFdufQZ R4VEzIZhDbaqUmBM2q+k=
3 new commits in eigen:
https://bitbucket.org/eigen/eigen/commits/4296e801a016/
Changeset: 4296e801a016
Branch: avx512-contraction-2
User: markdryan
Date: 2018-07-31 08:33:37+00:00
Summary: Fix tensor contraction for AVX512 machines
This patch modifies the TensorContraction class to ensure that the kc_ field is
always a multiple of the packet_size, if the packet_size is > 8. Without this
change spatial convolutions in Tensorflow do not work properly as the code that
re-arranges the input matrices can assert if kc_ is not a multiple of the
packet_size. This leads to a unit test failure,
//tensorflow/python/kernel_tests:conv_ops_test, on AVX512 builds of tensorflow.
Affected #: 1 file
https://bitbucket.org/eigen/eigen/commits/ad22199d184d/
Changeset: ad22199d184d
Branch: avx512-contraction-2
User: markdryan
Date: 2018-12-05 11:29:03+00:00
Summary: Fix evalShardedByInnerDim for AVX512 builds
evalShardedByInnerDim ensures that the values it passes for start_k and
end_k to evalGemmPartialWithoutOutputKernel are multiples of 8 as the kernel
does not work correctly when the values of k are not multiples of the
packet_size. While this precaution works for AVX builds, it is insufficient
for AVX512 builds where the maximum packet size is 16. The result is slightly
incorrect float32 contractions on AVX512 builds.
This commit fixes the problem by ensuring that k is always a multiple of
the packet_size if the packet_size is > 8.
Affected #: 1 file
https://bitbucket.org/eigen/eigen/commits/fb1d0e8cfa0f/
Changeset: fb1d0e8cfa0f
User: rmlarsen
Date: 2018-12-05 18:19:32+00:00
Summary: Merged in markdryan/eigen/avx512-contraction-2 (pull request #554)
Fix tensor contraction on AVX512 builds
Approved-by: Rasmus Munk Larsen <rmlarsen@xxxxxxxxxx>
Affected #: 2 files
Repository URL: https://bitbucket.org/eigen/eigen/
--
This is a commit notification from bitbucket.org. You are receiving
this because you have the service enabled, addressing the recipient of
this email.