[microNPU] Add NHWC -> NHCWB16 layout transformation pass #9561
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
15 tasks
ekalda
reviewed
Nov 23, 2021
There was a problem hiding this comment.
Looks really good! :) Very nice test coverage, clear and understandable tests and well done for checking all the various graph types! I noticed that none of the tests has a depthwise_conv2d in it, so maybe it is worth inserting a depthwise somewhere into the tested graphs as well to make sure no surprises in the depthwise front?
lhutton1
force-pushed
the
layout-optimize-pass-initial
branch
2 times, most recently
from
6c9a3d3 to
cf3b8da
Compare
lhutton1
requested review from
anijain2305,
areusch,
comaniac,
icemelon,
jroesch,
junrushao,
jwfromm,
MarisaKirisame,
mbrookhart,
merrymercy,
slyubomirsky,
tqchen,
vinx13,
wweic,
yzhliu,
zhiics and
ZihengJiang
as code owners
lhutton1
force-pushed
the
layout-optimize-pass-initial
branch
from
4e8527d to
3255971
Compare
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator. Change-Id: I0d6ad1f868dd6f78a236b4bc869e5cbd77c986b0
Change-Id: Iba3436d14a16106a696400c1329dcfa164631dc7
Change-Id: Ibdf1b129e11cb8e09b73cf73817532e6c8e8ee82
Change-Id: Ifc1b4c34cd891fce35d3e043ba81afdd2e34fc4e
Change-Id: I9d586c4931eb3499ed9c6c8555b1ff9c380f402b
lhutton1
force-pushed
the
layout-optimize-pass-initial
branch
from
3255971 to
ae84980
Compare
|
Friendly ping for comment/approval :) |
|
@ekalda ? |
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Jan 7, 2022
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
yangulei
pushed a commit
to yangulei/tvm
that referenced
this pull request
Jan 11, 2022
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
yangulei
pushed a commit
to yangulei/tvm
that referenced
this pull request
Jan 12, 2022
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Jan 13, 2022
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
qsqqsqqsq-intellif
pushed a commit
to qsqqsqqsq-intellif/tvm
that referenced
this pull request
Apr 29, 2022
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Adds a layout optimization pass that modifies the ifm/ofm layout of an operation to NHCWB16 where possible. This can occur when the producer or consumer of a tensor is also an NPU operator.
Note: this PR is dependent on #9560.cc @ekalda @manupa-arm @NicolaLancellotti @dchauhan-arm @mbaret