masahi commented on issue #8596:
URL: https://github.com/apache/tvm/issues/8596#issuecomment-1039909509
At one point can we expect a non-trivial demo?
--
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
huajsj commented on pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#issuecomment-1039901920
> Thanks @huajsj. This is interesting and tricky code, I'll try review
carefully. cc @comaniac
>
> Where does this PR sit in the feature road map?
Thanks @masahi ,
masahi commented on a change in pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#discussion_r806466392
##
File path: python/tvm/contrib/pipeline_executor.py
##
@@ -213,6 +209,16 @@ def get_output(self):
"""
return self._get_output()
+
masahi commented on a change in pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#discussion_r806466228
##
File path: python/tvm/contrib/pipeline_executor.py
##
@@ -213,6 +209,16 @@ def get_output(self):
"""
return self._get_output()
+
areusch commented on a change in pull request #46:
URL: https://github.com/apache/tvm-rfcs/pull/46#discussion_r806465372
##
File path: rfcs/0046-module-based-model-runtime-for-aot.md
##
@@ -0,0 +1,348 @@
+# Module-based Model Runtime Interface for AOT
+
+- Feature Name:
masahi edited a comment on pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#issuecomment-1039884305
Thanks @huajsj. This is interesting and tricky code, I'll try review
carefully. cc @comaniac
Where does this PR sit in the feature road map?
--
This is an
areusch commented on a change in pull request #46:
URL: https://github.com/apache/tvm-rfcs/pull/46#discussion_r806465372
##
File path: rfcs/0046-module-based-model-runtime-for-aot.md
##
@@ -0,0 +1,348 @@
+# Module-based Model Runtime Interface for AOT
+
+- Feature Name:
masahi commented on pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#issuecomment-1039884305
Thanks @huajsj. This is interesting and tricky code, I'll try review
carefully.
Where does this PR sit in the feature road map?
--
This is an automated message from the
areusch commented on a change in pull request #46:
URL: https://github.com/apache/tvm-rfcs/pull/46#discussion_r806455336
##
File path: rfcs/0046-module-based-model-runtime-for-aot.md
##
@@ -0,0 +1,348 @@
+# Module-based Model Runtime Interface for AOT
+
+- Feature Name:
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 0009a30 [TOPI] VNNI support for int8 dense (#10230)
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from b9fc7a8 [Tir]Adding detail error messages when MatchCopyPattern
function is failed. (#10244)
add a1d8f72 Fix a
masahi merged pull request #10230:
URL: https://github.com/apache/tvm/pull/10230
--
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:
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from e13110f Support sub warp reduction for CUDA target. (#10207)
add b9fc7a8 [Tir]Adding detail error messages when
masahi merged pull request #10245:
URL: https://github.com/apache/tvm/pull/10245
--
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:
masahi merged pull request #10244:
URL: https://github.com/apache/tvm/pull/10244
--
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:
masahi commented on pull request #10229:
URL: https://github.com/apache/tvm/pull/10229#issuecomment-1039822161
@Lunderberg Please take a look at the CI problem.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
This is an automated email from the ASF dual-hosted git repository.
wuwei 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 e13110f Support sub warp reduction for CUDA target.
vinx13 merged pull request #10207:
URL: https://github.com/apache/tvm/pull/10207
--
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:
masahi commented on a change in pull request #10248:
URL: https://github.com/apache/tvm/pull/10248#discussion_r806415906
##
File path: src/relay/transforms/device_planner.cc
##
@@ -22,9 +22,6 @@
* \brief Determines a unique \p VirtualDevice to hold the result of every
Relay
manupa-arm commented on a change in pull request #8509:
URL: https://github.com/apache/tvm/pull/8509#discussion_r806407862
##
File path: include/tvm/ir/module.h
##
@@ -349,6 +386,9 @@ class IRModuleNode : public Object {
*/
std::unordered_set import_set_;
friend
masahi edited a comment on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039788394
Are you sure you can run your PT model on cuda via cudnn (use nvprof)? I'm
getting `CUDNN_STATUS_BAD_PARAM` when trying to run your model with cudnn. As
always, cuDNN error
masahi commented on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039788394
Are you sure you can run your PT model on cuda via cudnn? I'm getting
`CUDNN_STATUS_BAD_PARAM` when trying to run your model with cudnn. A WIP branch
junrushao1994 commented on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1039779883
will leave the PR to @vinx13 and @masahi for a second look :-)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
This is an automated email from the ASF dual-hosted git repository.
junrushao pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from be17697 clean up conv2d type rels (#10236)
add de73b99 [Torch] Fix conv2d transpose with group (#10235)
No new
junrushao1994 merged pull request #10235:
URL: https://github.com/apache/tvm/pull/10235
--
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:
junrushao1994 closed issue #10223:
URL: https://github.com/apache/tvm/issues/10223
--
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:
junrushao1994 commented on pull request #10235:
URL: https://github.com/apache/tvm/pull/10235#issuecomment-1039759287
right. I noticed the same issue very recently. Thanks for fixing this!!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
This is an automated email from the ASF dual-hosted git repository.
jroesch pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 2a840a3 [Hexagon] Refactor Hexagon.cmake (#10227)
add be17697 clean up conv2d type rels (#10236)
No new
jroesch merged pull request #10236:
URL: https://github.com/apache/tvm/pull/10236
--
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:
vinx13 commented on a change in pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#discussion_r805065309
##
File path: tests/python/unittest/test_tir_transform_instrument_bound_checkers.py
##
@@ -94,7 +94,7 @@ def test_out_of_bounds_vectorize_llvm(nn, index_a,
driazati opened a new pull request #10247:
URL: https://github.com/apache/tvm/pull/10247
This updates the docker image build to use Python 3.7 and Ubuntu 18 as
discussed in #9703. The update is mostly straightforward except that the `apt`
boost isn't built with 3.7 so we must now build it
driazati opened a new pull request #10246:
URL: https://github.com/apache/tvm/pull/10246
This sets the builds for each platform to use
[sccache](https://github.com/mozilla/sccache) as a build cache. For a fully
cached rebuild, this reduces the time (in testing) from 30+ minutes to 1-2
huajsj opened a new pull request #10245:
URL: https://github.com/apache/tvm/pull/10245
lint.sh complain for an addtional space line in 'utils.cc', just fix it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
huajsj opened a new pull request #10244:
URL: https://github.com/apache/tvm/pull/10244
There is an error message to show the body when 'MatchCopyPattern' is
failed, but the error message not give the information why this function get
failed. Adding the detail error information to help
masahi commented on pull request #10230:
URL: https://github.com/apache/tvm/pull/10230#issuecomment-1039588965
Another perf result, this time on a desktop CPU `i5-11400 @ 2.60GHz`, 6
threads.
TVM showing excellent performance!
M|N|K|TVM|FBGEMM
-- | -- | -- | --| -- |
masahi commented on pull request #10239:
URL: https://github.com/apache/tvm/pull/10239#issuecomment-1039572343
Please also show the reference result (accuracy without fq2i).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
masahi commented on pull request #10239:
URL: https://github.com/apache/tvm/pull/10239#issuecomment-1039484466
Can you show some accuracy results on QAT models (BERT etc)?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
Tantalus13A98B5F commented on pull request #8605:
URL: https://github.com/apache/tvm/pull/8605#issuecomment-1039458148
Hi @mengbingrock
Not sure how long have you been going, but `bsr_conv2d.convert2` is merely
for graph & param transformation, and it assumes well-padded shapes (0 for
srkreddy1238 opened a new pull request #10243:
URL: https://github.com/apache/tvm/pull/10243
Thanks for contributing to TVM! Please refer to guideline
https://tvm.apache.org/docs/contribute/ for useful information and tips. After
the pull request is submitted, please request code
huajsj commented on pull request #10234:
URL: https://github.com/apache/tvm/pull/10234#issuecomment-1039411862
@comaniac , @masahi , please take a look.
--
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
jacobbohlin commented on pull request #10242:
URL: https://github.com/apache/tvm/pull/10242#issuecomment-1039254125
@manupa-arm @mbaret @lhutton1 @ekalda @NicolaLancellotti @dchauhan-arm
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
jacobbohlin opened a new pull request #10242:
URL: https://github.com/apache/tvm/pull/10242
This adds MobileNetv2 tflite testcase to the microNPU's network tests.
Co-authored-by: Jacob Bohlin
--
This is an automated message from the Apache Git Service.
To respond to the message,
JCBrouwer commented on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039250272
Yes using CUDNN is fine with me, that would be great :)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
masahi edited a comment on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039219766
Yeah, cuda backend doesnt support groups. We should fix that but I am not
looking to do it. A PR welcome.
You can try the cpu backend to verify the result. If you are ok
masahi commented on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039219766
Yeah, cuda backend doesnt support groups. You can try the cpu backend to
verify the result. If you are ok with using cudnn, I can quickly enable groups
support for cudnn conv
spanijel opened a new pull request #10241:
URL: https://github.com/apache/tvm/pull/10241
Adding compiler support for TANH operator, which is based on
an underlying pattern matching scheme.
One negative test is included as well.
cc @leandron @Leo-arm @mbaret @u99127 for
tqchen commented on a change in pull request #9748:
URL: https://github.com/apache/tvm/pull/9748#discussion_r805954346
##
File path: include/tvm/runtime/container/optional.h
##
@@ -93,6 +93,11 @@ class Optional : public ObjectRef {
ICHECK(data_ != nullptr);
return
pfk-beta opened a new issue #10240:
URL: https://github.com/apache/tvm/issues/10240
Hello,
I'm using heavily RPC tracker and server with TVM. I encounter many times
issue, when tuning task cause device reboot or rpc_android_app restart. After
that, whole tuning process need to be
This is an automated email from the ASF dual-hosted git repository.
kparzysz pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from ce45f26 add a simplify rule for floordiv(x*8+7, 16) => floordiv(x, 2)
(#10232)
add 2a840a3 [Hexagon] Refactor
This is an automated email from the ASF dual-hosted git repository.
kparzysz pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from ac9bf1f [TVMC] Add configuration `tir.add_lower_pass` to option
`--pass-config` (#9817)
add ce45f26 add a
pfk-beta closed issue #8732:
URL: https://github.com/apache/tvm/issues/8732
--
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:
pfk-beta commented on issue #8732:
URL: https://github.com/apache/tvm/issues/8732#issuecomment-1039155944
I think that I was using old version of repository. In second try I have
working OpelCL runtime with RPC server/tracker and android app, so I'm closing
issue.
--
This is an
kparzysz-quic merged pull request #10227:
URL: https://github.com/apache/tvm/pull/10227
--
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:
pfk-beta closed issue #10202:
URL: https://github.com/apache/tvm/issues/10202
--
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:
pfk-beta commented on issue #10202:
URL: https://github.com/apache/tvm/issues/10202#issuecomment-1039152752
Issue should be solved by PR https://github.com/apache/tvm/pull/10227 . I'm
closing issue as the issue author:)
--
This is an automated message from the Apache Git Service.
To
kparzysz-quic merged pull request #10232:
URL: https://github.com/apache/tvm/pull/10232
--
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:
mengbingrock edited a comment on pull request #8605:
URL: https://github.com/apache/tvm/pull/8605#issuecomment-1039109093
Hi @Tantalus13A98B5F
Your work is very helpful!
I'm evaluating the performance of sparse kernels. I noticed that PR
currently only supports padding = 1 case if
This is an automated email from the ASF dual-hosted git repository.
leandron 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 ac9bf1f [TVMC] Add configuration
leandron commented on pull request #9817:
URL: https://github.com/apache/tvm/pull/9817#issuecomment-1039135305
Thanks @leeexyz, this is merged now.
--
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
leandron merged pull request #9817:
URL: https://github.com/apache/tvm/pull/9817
--
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:
mengbingrock commented on pull request #8605:
URL: https://github.com/apache/tvm/pull/8605#issuecomment-1039109093
Hi @Tantalus13A98B5F
Your work is very helpful!
I'm evaluating the performance of sparse kernels. I noticed that PR
currently only supports padding = 1 case if I'm
JCBrouwer commented on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1039101873
Great, thanks for the quick response!
I can confirm that it's converting from PyTorch correctly for my more
complex model on your branch.
Now I'm getting an
leandron commented on issue #10238:
URL: https://github.com/apache/tvm/issues/10238#issuecomment-1038923382
Me and @grant-arm did some investigation in this and we found out that the
fixed/safe versions of Pillow don't release for Python 3.6 anymore.
So this is one more reason for
manupa-arm commented on issue #10238:
URL: https://github.com/apache/tvm/issues/10238#issuecomment-1038884564
cc : @grant-arm
--
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
This is an automated email from the ASF dual-hosted git repository.
mousius pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from bb60ee9 [PyTorch] add var_mean support (#10233)
add 55849e6 [USMP] adding support for U2 and U3 usecases (#10193)
Mousius commented on pull request #10193:
URL: https://github.com/apache/tvm/pull/10193#issuecomment-1038879556
Thanks @manupa-arm :smile_cat:
--
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
Mousius merged pull request #10193:
URL: https://github.com/apache/tvm/pull/10193
--
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:
67 matches
Mail list logo