tvm-bot commented on PR #14303:
URL: https://github.com/apache/tvm/pull/14303#issuecomment-1469387339
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
multiverstack-intellif opened a new pull request, #14303:
URL: https://github.com/apache/tvm/pull/14303
Add llvm-15 and mlir-15 for RFC: [Introduce
PresburgerSet](https://github.com/apache/tvm-rfcs/pull/99)
--
This is an automated message from the Apache Git Service.
To respond to the mes
This is an automated email from the ASF dual-hosted git repository.
echuraev pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
from 970cd1def8 [TIR][Hexagon] Enhancement of NarrowDataType pass for
binary ops (#14298)
add 6eb4b873e1 [DOCS][ADRENO]
echuraev merged PR #13867:
URL: https://github.com/apache/tvm/pull/13867
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.
echuraev commented on PR #14299:
URL: https://github.com/apache/tvm/pull/14299#issuecomment-1469371606
Thank you for the PR. It adds nice functionality. Quickly took a look at the
changes and left several comments.
--
This is an automated message from the Apache Git Service.
To respond to
echuraev commented on code in PR #14299:
URL: https://github.com/apache/tvm/pull/14299#discussion_r1136520765
##
3rdparty/cutlass:
##
Review Comment:
Did you change something in this submodule?
##
adreno_recording/adreno_recording_executor.py:
##
@@ -0,0
This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch nightly
in repository https://gitbox.apache.org/repos/asf/tvm.git
from ccc0b9162f [fix][relay][qnn] Bug fix for 8-bit quantized mul (#14286)
add c9ab1979f1 [CI][ETHOSN] Add ssh to th
masahi merged PR #14298:
URL: https://github.com/apache/tvm/pull/14298
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.or
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new 970cd1def8 [TIR][Hexagon] Enhancement of NarrowDataType
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new 075e2ec7bb [Frontend][Paddle]fix eye and dist (#14292)
masahi merged PR #14292:
URL: https://github.com/apache/tvm/pull/14292
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.or
This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new e22a2d5b9f [IR] Enhance IRModule SEqual/SHash to suppor
tqchen merged PR #14289:
URL: https://github.com/apache/tvm/pull/14289
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.or
sunggg commented on code in PR #14282:
URL: https://github.com/apache/tvm/pull/14282#discussion_r1136411192
##
src/runtime/relax_vm/builtin.cc:
##
@@ -380,6 +380,40 @@
TVM_REGISTER_GLOBAL("vm.builtin.make_tuple").set_body([](TVMArgs args, TVMRetVal
*rv = arr;
});
+TVM_REG
driazati commented on issue #14260:
URL: https://github.com/apache/tvm/issues/14260#issuecomment-1469058137
+1
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscrib
sunggg commented on code in PR #14282:
URL: https://github.com/apache/tvm/pull/14282#discussion_r1136393998
##
src/relax/op/op.cc:
##
@@ -315,6 +315,32 @@ Expr MakeShapeOf(Expr expr) {
TVM_REGISTER_GLOBAL("relax.op.shape_of").set_body_typed(MakeShapeOf);
+// tensor_to_shape
sunggg commented on code in PR #14282:
URL: https://github.com/apache/tvm/pull/14282#discussion_r1136385936
##
tests/python/relax/test_transform_legalize_ops_manipulate.py:
##
@@ -537,6 +584,54 @@ def reshape(var_rxplaceholder: T.handle, var_T_reshape:
T.handle):
mod = Leg
sunggg commented on code in PR #14282:
URL: https://github.com/apache/tvm/pull/14282#discussion_r1136385936
##
tests/python/relax/test_transform_legalize_ops_manipulate.py:
##
@@ -537,6 +584,54 @@ def reshape(var_rxplaceholder: T.handle, var_T_reshape:
T.handle):
mod = Leg
tqchen opened a new pull request, #14302:
URL: https://github.com/apache/tvm/pull/14302
This PR adds an instrumentation option to the relax VM. The instrument will
be called before/after each call instruction if specified.
We also include a testing utility that leverages uses instrume
tvm-bot commented on PR #14302:
URL: https://github.com/apache/tvm/pull/14302#issuecomment-1469028576
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
This is an automated email from the ASF dual-hosted git repository.
areusch pushed a commit to branch v0.11.0
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/v0.11.0 by this push:
new 046910a410 [Release] Adjust version to 0.11.1 (#
areusch commented on issue #14260:
URL: https://github.com/apache/tvm/issues/14260#issuecomment-1468974281
+1
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe
areusch merged PR #14300:
URL: https://github.com/apache/tvm/pull/14300
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.o
areusch commented on issue #14260:
URL: https://github.com/apache/tvm/issues/14260#issuecomment-1468882801
@leandron i think that PR should unblock the release
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL ab
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new ce1fa8908f [TE] Record primitives of Schedule for visua
masahi merged PR #14168:
URL: https://github.com/apache/tvm/pull/14168
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.or
leandron commented on PR #14300:
URL: https://github.com/apache/tvm/pull/14300#issuecomment-1468803958
There was some confusion from my side to get the commit signed and verified
by GitHub, now it seems all is fixed. Thanks @areusch
--
This is an automated message from the Apache Git Ser
masahi commented on PR #14293:
URL: https://github.com/apache/tvm/pull/14293#issuecomment-1468781911
Do you mean this https://github.com/tlc-pack/tlcpack? I'm not familiar with
this package, but its README says how to build an image with a specific CUDA
version. Is there other action requir
masahi commented on code in PR #14215:
URL: https://github.com/apache/tvm/pull/14215#discussion_r1136138654
##
tests/python/relax/test_transform_alter_op_impl.py:
##
@@ -0,0 +1,342 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license ag
PhilippvK commented on PR #14272:
URL: https://github.com/apache/tvm/pull/14272#issuecomment-1468690703
> One way that I can think of, although it's not very elegant, is to allow a
special nn.conv2d_depthwise parameter thorough desired_layout_ops, then
override the FTVMConvertOpLayout funct
ibsidorenko commented on PR #14298:
URL: https://github.com/apache/tvm/pull/14298#issuecomment-1468687086
@tvm-bot rerun
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To
tvm-bot commented on PR #14301:
URL: https://github.com/apache/tvm/pull/14301#issuecomment-1468686407
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
driazati opened a new pull request, #14301:
URL: https://github.com/apache/tvm/pull/14301
This updates the docs with some of the discussion and methods from
the v0.11.1 release issue #14260.
--
This is an automated message from the Apache Git Service.
To respond to the message, please l
AndrewZhaoLuo commented on code in PR #14167:
URL: https://github.com/apache/tvm/pull/14167#discussion_r1136006711
##
python/tvm/meta_schedule/schedule/cuda/layout_transform.py:
##
@@ -0,0 +1,321 @@
+import tvm
+from tvm import topi
+import math
+from typing import List, Sequenc
AndrewZhaoLuo commented on code in PR #14167:
URL: https://github.com/apache/tvm/pull/14167#discussion_r1136006711
##
python/tvm/meta_schedule/schedule/cuda/layout_transform.py:
##
@@ -0,0 +1,321 @@
+import tvm
+from tvm import topi
+import math
+from typing import List, Sequenc
Mousius commented on PR #14179:
URL: https://github.com/apache/tvm/pull/14179#issuecomment-1468569866
As with other compilers, ahead of time compilation can be applied to a much
broader scope than just embedded and the graph executor does work in more
constrained environments, is there any
This is an automated email from the ASF dual-hosted git repository.
ruihangl pushed a commit to branch unity
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/unity by this push:
new 3c65ac467d [Unity][Op] Cumsum (#14297)
3c65ac467d i
MasterJH5574 merged PR #14297:
URL: https://github.com/apache/tvm/pull/14297
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apa
leandron commented on issue #14260:
URL: https://github.com/apache/tvm/issues/14260#issuecomment-1468567540
I submitted https://github.com/apache/tvm/pull/14300 to deal with the
release updates. Am I missing other places that need updating in `apache/tvm`
repo?
--
This is an automated me
leandron opened a new pull request, #14300:
URL: https://github.com/apache/tvm/pull/14300
* Adjust version numbers for source package release, which does not contain
git metadata to infer the release version.
cc @areusch @driazati @Johnson9009 for reviews
--
This is an automated me
tvm-bot commented on PR #14300:
URL: https://github.com/apache/tvm/pull/14300#issuecomment-1468565981
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
areusch commented on PR #14179:
URL: https://github.com/apache/tvm/pull/14179#issuecomment-1468482950
@Mousius i think it's two angles to the same problem: GraphExecutor isn't a
great choice with the C runtime. However, you can in practice use AOT with the
C++ runtime. I would think in term
leandron commented on issue #14260:
URL: https://github.com/apache/tvm/issues/14260#issuecomment-1468468263
OK, so I'm progressing with option 1 for the existing `v0.11.1` branch. We
can review these rules for the next releases.
--
This is an automated message from the Apache Git Service.
This is an automated email from the ASF dual-hosted git repository.
leandron pushed a change to branch docker_llvm_15_update
in repository https://gitbox.apache.org/repos/asf/tvm.git
omit e692ae39f2 Update ci_arm to have LLVM 15
add 895f55df6a Update ci_arm to have LLVM 15
This update
lhutton1 commented on code in PR #14272:
URL: https://github.com/apache/tvm/pull/14272#discussion_r1135821786
##
python/tvm/driver/tvmc/transform.py:
##
@@ -210,9 +229,8 @@ def generate_transform_args(parser):
# AlterLayout
parser.add_argument(
"--desired-layo
slyubomirsky commented on issue #14287:
URL: https://github.com/apache/tvm/issues/14287#issuecomment-1468439077
Conclusion from Relax open development meeting: We think this error is
correct and that we should not permit `DataflowVar`s in `StructInfo`
annotations. Using an explicit `R.outpu
slyubomirsky closed issue #14287: [Unity][Discuss] Scoping rule for StructInfo
URL: https://github.com/apache/tvm/issues/14287
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
T
yongwww commented on issue #14287:
URL: https://github.com/apache/tvm/issues/14287#issuecomment-1468388640
`print(Mod["main"].body.body.struct_info_) `# gv's struct_info_
out: `R.Tensor(lv, dtype="float32”) `
--
This is an automated message from the Apache Git Service.
To respon
neildhickey commented on PR #14296:
URL: https://github.com/apache/tvm/pull/14296#issuecomment-1468387506
LGTM!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscri
ekalda commented on PR #14175:
URL: https://github.com/apache/tvm/pull/14175#issuecomment-1468364272
Sorry this fell off my radar for a bit - updated the patch now with a test
to check that we set the `logging.basicConfig`.
--
This is an automated message from the Apache Git Service.
To
ekalda commented on code in PR #14175:
URL: https://github.com/apache/tvm/pull/14175#discussion_r1135777206
##
tests/python/driver/tvmc/test_command_line.py:
##
@@ -210,3 +210,45 @@ def test_tvmc_compile_input_model(mock_compile_model,
tmpdir_factory, model):
_main(run_arg
tvm-bot commented on PR #14299:
URL: https://github.com/apache/tvm/pull/14299#issuecomment-1468277687
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
valmat07 opened a new pull request, #14299:
URL: https://github.com/apache/tvm/pull/14299
This PR introduces the use of write queues for OpenCL
For this, a separate class was implemented for graph_executor, which
contains functions for recording and running what has already been recor
This is an automated email from the ASF dual-hosted git repository.
lukhut pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new d22bdce2bf [Relay][Op] Connect existing arm_cpu schedul
lhutton1 merged PR #14270:
URL: https://github.com/apache/tvm/pull/14270
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.
ibsidorenko opened a new pull request, #14298:
URL: https://github.com/apache/tvm/pull/14298
This is enhancement of [PR#13327](https://github.com/apache/tvm/pull/13327).
**Motivation**:
Playing with MetaScheduler for Hexagon target it was found that `avg_pool2d`
has rather poor per
tvm-bot commented on PR #14298:
URL: https://github.com/apache/tvm/pull/14298#issuecomment-1468101193
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
tvm-bot commented on PR #14297:
URL: https://github.com/apache/tvm/pull/14297#issuecomment-1468086502
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
SiriusNEO opened a new pull request, #14297:
URL: https://github.com/apache/tvm/pull/14297
This PR introduces the cumulative sum high-level operator.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go
psrivas2 commented on code in PR #14215:
URL: https://github.com/apache/tvm/pull/14215#discussion_r1135459707
##
tests/python/relax/test_transform_alter_op_impl.py:
##
@@ -0,0 +1,342 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license
leandron commented on PR #14293:
URL: https://github.com/apache/tvm/pull/14293#issuecomment-1467858897
My experience with CUDA and the whole CUDA ecosystem is very limited, but
just wanted to point out that it would be good to review also the tlcpack CUDA
versions, perhaps to propose this n
srkreddy1238 commented on code in PR #13867:
URL: https://github.com/apache/tvm/pull/13867#discussion_r1135342636
##
tests/scripts/setup-adreno-env.sh:
##
@@ -0,0 +1,113 @@
+#!/usr/bin/env bash
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor
leandron commented on PR #14295:
URL: https://github.com/apache/tvm/pull/14295#issuecomment-1467835539
#14296
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe
tvm-bot commented on PR #14296:
URL: https://github.com/apache/tvm/pull/14296#issuecomment-1467834270
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
leandron opened a new pull request, #14296:
URL: https://github.com/apache/tvm/pull/14296
Update `ci_arm` to have support for LLVM 15, which brings better support for
vectorisation.
The proposed docker tag is: `20230314-060145-ccc0b9162`.
Note: I'll run a CI run to test the
This is an automated email from the ASF dual-hosted git repository.
leandron pushed a change to branch docker_llvm_15_update
in repository https://gitbox.apache.org/repos/asf/tvm.git
at e692ae39f2 Update ci_arm to have LLVM 15
No new revisions were added by this update.
leandron commented on PR #14295:
URL: https://github.com/apache/tvm/pull/14295#issuecomment-1467832321
I'll reopen as a branch of `apache/tvm`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the
leandron closed pull request #14295: [Do not merge][CI] Update ci_arm docker
image to have LLVM 15
URL: https://github.com/apache/tvm/pull/14295
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the s
leandron opened a new pull request, #14295:
URL: https://github.com/apache/tvm/pull/14295
Update `ci_arm` to have support for LLVM 15, which brings better support for
vectorisation.
The proposed docker tag is: `20230314-060145-ccc0b9162`.
Note: I'll run a CI run to test the
tvm-bot commented on PR #14295:
URL: https://github.com/apache/tvm/pull/14295#issuecomment-1467829870
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
This is an automated email from the ASF dual-hosted git repository.
ashutoshp pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
from ccc0b9162f [fix][relay][qnn] Bug fix for 8-bit quantized mul (#14286)
add c9ab1979f1 [CI][ETHOSN] Add ssh to the dr
ashutosh-arm merged PR #14246:
URL: https://github.com/apache/tvm/pull/14246
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apa
ashutosh-arm commented on PR #14246:
URL: https://github.com/apache/tvm/pull/14246#issuecomment-1467814246
Submitting this now. We can take up other changes in a follow up if needed.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to Gi
apeskov commented on PR #14231:
URL: https://github.com/apache/tvm/pull/14231#issuecomment-1467791666
I guess this PR was submitted by mistake. Looks like this dev branch has
some minimal PoC with offloading SoftMax to OneDNN for GPU device. It's
definitely not ready for merging.
Cou
lileidev commented on issue #14137:
URL: https://github.com/apache/tvm/issues/14137#issuecomment-1467787359
Both Module1 and Module0 can run pass on my machine.
Module1:
2023-03-14 10:05:44 [INFO] [task_scheduler.cc:260] Task #0 has finished.
Remaining task(s): 0
2023-03-14 10:05:
quic-sanirudh commented on PR #14294:
URL: https://github.com/apache/tvm/pull/14294#issuecomment-1467780883
cc @Lunderberg @kparzysz-quic
There is a comment in
[`layout_transformation.cc`](https://github.com/apache/tvm/blob/ccc0b9162f2e983a8810e99c903c7141dbec81b6/src/tir/schedule/pri
tvm-bot commented on PR #14294:
URL: https://github.com/apache/tvm/pull/14294#issuecomment-1467772001
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
quic-sanirudh opened a new pull request, #14294:
URL: https://github.com/apache/tvm/pull/14294
We're adding support to generate `llvm.assume` from `tir.assume` as we
currently see an error when lowering code with tir.assume in certain cases.
--
This is an automated message from the Apache
tvm-bot commented on PR #14293:
URL: https://github.com/apache/tvm/pull/14293#issuecomment-1467767990
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
masahi opened a new pull request, #14293:
URL: https://github.com/apache/tvm/pull/14293
The last time we updated the CUDA version on CI was [two years
ago](https://github.com/apache/tvm/pull/8119), I think it is a good time for
another update.
v11.7 is the latest version officially s
masahi commented on code in PR #14291:
URL: https://github.com/apache/tvm/pull/14291#discussion_r1135119116
##
src/runtime/contrib/cublas/cublas.cc:
##
@@ -133,6 +134,120 @@ bool CheckMixPrecisionType(DLDataType in_dtype,
DLDataType out_dtype, bool int_s
int roundoff(int v, in
tvm-bot commented on PR #14292:
URL: https://github.com/apache/tvm/pull/14292#issuecomment-1467564531
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
XG-zheng opened a new pull request, #14292:
URL: https://github.com/apache/tvm/pull/14292
Remove redundant `abs` op for `dist`.
Cover the case where `num_columns` is equal -1 for `eye`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log o
masahi opened a new pull request, #14291:
URL: https://github.com/apache/tvm/pull/14291
To support dynamic shape better than cutlass BYOC
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the speci
tvm-bot commented on PR #14291:
URL: https://github.com/apache/tvm/pull/14291#issuecomment-1467518614
Thanks for contributing to TVM! Please refer to the contributing guidelines
https://tvm.apache.org/docs/contribute/ for useful information and tips. Please
request code reviews from
85 matches
Mail list logo