MarisaKirisame commented on a change in pull request #7083:
URL: https://github.com/apache/tvm/pull/7083#discussion_r553767924
##
File path: src/relay/op/random/kernel.cc
##
@@ -0,0 +1,89 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more
insop commented on issue #7186:
URL: https://github.com/apache/tvm/issues/7186#issuecomment-756578686
For `_npi_advanced_indexing_multiple ` is at
https://github.com/apache/tvm/pull/7230#pullrequestreview-564031534, PTAL.
insop commented on a change in pull request #7230:
URL: https://github.com/apache/tvm/pull/7230#discussion_r553766990
##
File path: tests/python/frontend/mxnet/test_forward.py
##
@@ -1935,6 +1935,29 @@ def verify(data_shape, axis, use_length, length):
verify((2, 3, 4), 2,
insop opened a new pull request #7230:
URL: https://github.com/apache/tvm/pull/7230
Adding _npi_advanced_indexing_multiple as discussed in
https://github.com/apache/tvm/issues/7186
Need to find a proper ` mx.sym.np` and wanted ask for the reviewers
help to find it, so that
Meteorix commented on pull request #7146:
URL: https://github.com/apache/tvm/pull/7146#issuecomment-756577894
@jcf94 @merrymercy @Laurawly finally the ci passed. Also I have fixed the
dtype check for batch_matmul. Please review this mr again.
masahi commented on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756541283
Thanks @mbrookhart
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.
masahi pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 4911a08 [BUGFIX] Change debug_runtime to represent times in seconds
internally (#7227)
add 4e8cc4f Fix Get Valid
masahi merged pull request #7229:
URL: https://github.com/apache/tvm/pull/7229
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
masahi edited a comment on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756513178
This is not important, but I'd suggest swapping the outer most `then block`
and `else block` (`with ib.if_scope(num_anchors == 0):`), since the else block
is so far apart and
masahi edited a comment on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756513178
This is not important, but I'd suggest swapping the outer most `then block`
and `else block` (`with ib.if_scope(num_anchors == 0):`), since the else block
is so far apart and
masahi edited a comment on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756513178
This is not important, but I'd suggest swapping the outer most `then block`
and `else block` (`with ib.if_scope(num_anchors == 0):`), since the else block
is so far apart and
masahi edited a comment on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756513178
This is not important, but I'd suggest swapping the outer most `then block`
and `else block` (`with ib.if_scope(num_anchors == 0):`), since the else block
is so far apart and
masahi commented on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756513178
This is not important, but I'd suggest swapping the outer most `then block`
and `else block` (`with ib.if_scope(num_anchors == 0):`), since the else block
is so far apart and hard
comaniac commented on a change in pull request #7226:
URL: https://github.com/apache/tvm/pull/7226#discussion_r553691384
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -315,11 +315,10 @@ def auto_schedule_topi(outs):
new_key =
jcf94 commented on a change in pull request #7226:
URL: https://github.com/apache/tvm/pull/7226#discussion_r553690488
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -315,11 +315,10 @@ def auto_schedule_topi(outs):
new_key =
mbrookhart commented on pull request #7229:
URL: https://github.com/apache/tvm/pull/7229#issuecomment-756490687
cc @masahi
This is an automated message from the Apache Git Service.
To respond to the message, please log on
mbrookhart opened a new pull request #7229:
URL: https://github.com/apache/tvm/pull/7229
Thanks @trevor-m and @anijain2305 for finding this issue. Sometimes a model
may propose 0 boxes, and the code had a place that ended up assigning
uninitialized memory to valid_counts, which could
This is an automated email from the ASF dual-hosted git repository.
jcf94 pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 8910f72 [ConvertLayout] Support transpose (#7214)
add 4911a08 [BUGFIX] Change debug_runtime to represent times in
jcf94 merged pull request #7227:
URL: https://github.com/apache/tvm/pull/7227
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
jcf94 closed issue #7207:
URL: https://github.com/apache/tvm/issues/7207
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
jcf94 commented on pull request #7227:
URL: https://github.com/apache/tvm/pull/7227#issuecomment-756487144
Thanks!
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
tqchen opened a new pull request #7228:
URL: https://github.com/apache/tvm/pull/7228
cc @tmoreau89 @junrushao1994 @comaniac
This is an automated message from the Apache Git Service.
To respond to the message, please log on
comaniac commented on pull request #7214:
URL: https://github.com/apache/tvm/pull/7214#issuecomment-756439217
Thanks @anijain2305 @jcf94
This is an automated message from the Apache Git Service.
To respond to the message,
comaniac merged pull request #7214:
URL: https://github.com/apache/tvm/pull/7214
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
This is an automated email from the ASF dual-hosted git repository.
comaniac 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 8910f72 [ConvertLayout] Support transpose (#7214)
areusch commented on pull request #7227:
URL: https://github.com/apache/tvm/pull/7227#issuecomment-756419891
@tqchen @junrushao1994 @FrozenGene
This is an automated message from the Apache Git Service.
To respond to the
areusch opened a new pull request #7227:
URL: https://github.com/apache/tvm/pull/7227
Follow-on to #7197 and #6964, and fix #7207.
DebugRuntime used to represent runtimes as doubles but in microseconds. An
unintentionally merged part of #6964 migrated the timing parts to use
tmoreau89 commented on pull request #7083:
URL: https://github.com/apache/tvm/pull/7083#issuecomment-756343478
Thanks for the reviews, @junrushao1994 @MarisaKirisame @eric-haibin-lin if
you have some cycles, input would be appreciated!
trevor-m commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756341660
> Thanks, I'll try to reproduce. You're building with thrust and TRT, right?
>
> you can't compile a cuda kernel with zero threads, so we always make sure
it's at least 1:
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756340257
Thanks, I'll try to reproduce. You're building with thrust and TRT, right?
you can't compile a cuda kernel with zero threads, so we always make sure
it's at least 1:
trevor-m commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756338926
Thanks @mbrookhart , I tried that but it didn't fix the error.
Here is a script to reproduce the error:
https://gist.github.com/trevor-m/f44d3d0e7edcaee12722e518e5959b82
anijain2305 commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756337655
> Alas. I would still very much appreciate the script to reproduce this so I
can hunt it down.
Yes, Trevor is working on it. It needs TRT workflow and thats why the
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756334561
https://github.com/apache/tvm/blob/9815ae2d9e17eece1a1009eb6436c80f931c734e/python/tvm/topi/cuda/nms.py#L177-L179
Log(0) is undefined, we should probably just wrap the entire
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756333652
Alas. I would still very much appreciate the script to reproduce this so I
can hunt it down.
This is an
anijain2305 edited a comment on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756327373
> I don't this this is valid if num_anchors is zero, it could lead to
undefined behavior. Could you wrap that in an `with ib.if_scope(num_anchors >
0)` and see if that
anijain2305 commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756327373
> I don't this this is valid if num_anchors is zero, it could lead to
undefined behavior. Could you wrap that in an `with ib.if_scope(num_anchors >
0)` and see if that fixes
comaniac opened a new pull request #7226:
URL: https://github.com/apache/tvm/pull/7226
We currently return an initial (naive) schedule in the tracing mode due to
no tuned schedules. However, this won't work in the JIT mode, as the function
will be compiled immediately after lowering. This
areusch commented on issue #7207:
URL: https://github.com/apache/tvm/issues/7207#issuecomment-756305889
no, it was tangential to that PR, so this issue is open to track a follow-on
to fix
This is an automated message from
areusch commented on pull request #7225:
URL: https://github.com/apache/tvm/pull/7225#issuecomment-756305451
@tom-gall want to add discovery board to the [list of supported
boards](https://github.com/apache/tvm/blob/main/docs/microtvm/index.rst#supported-hardware)?
in practice I think
mbrookhart edited a comment on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756298093
I don't this this is valid if num_anchors is zero, it could lead to
undefined behavior. Could you wrap that in an `with ib.if_scope(num_anchors >
0)` and see if that
junrushao1994 commented on issue #7207:
URL: https://github.com/apache/tvm/issues/7207#issuecomment-756298316
I suppose the bug has been fixed by #7197, is that correct?
This is an automated message from the Apache Git
mbrookhart edited a comment on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756298093
I don't this this is valid if num_anchors is zero, it could lead to
undefined behavior. Could you wrap that in an `with ib.if_scope(num_anchors >
0)`? and see if that
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756298093
I don't this this is valid if num_anchors is zero, it could lead to
undefined behavior. Could you wrap that in an `with ib.if_scope(num_anchors >
0)`? and see if that fixes the
tom-gall commented on pull request #7225:
URL: https://github.com/apache/tvm/pull/7225#issuecomment-756289695
@areusch @gromero
This is an automated message from the Apache Git Service.
To respond to the message, please log
tom-gall opened a new pull request #7225:
URL: https://github.com/apache/tvm/pull/7225
This one line patch adds support for the ST STM32F746 Discovery board with
µTVM.
Signed-off-by: Tom Gall
This is an
junrushao1994 commented on a change in pull request #7153:
URL: https://github.com/apache/tvm/pull/7153#discussion_r553499647
##
File path: cmake/modules/Libbacktrace.cmake
##
@@ -0,0 +1,18 @@
+include(ExternalProject)
+
+ExternalProject_Add(project_libbacktrace
+ PREFIX
junrushao1994 closed issue #7204:
URL: https://github.com/apache/tvm/issues/7204
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
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756278828
Ooh, interesting, doing NMS on no boxes, I'll take a look with that idea.
This is an automated message from the
trevor-m edited a comment on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756277499
Thanks for looking into it and finding that info @mbrookhart !
Here is the relevant relay graph:
```
boxes = relay.var("boxes", shape=(1, relay.Any(), 5),
trevor-m commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756277499
Thanks for the info @mbrookhart !
Here is the relevant relay graph:
```
boxes = relay.var("boxes", shape=(1, relay.Any(), 5), dtype="float32")
max_output_size =
mbrookhart edited a comment on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756239725
Looking at the code, assuming you have thrust enabled, this should be
kernel0:
mbrookhart commented on pull request #7123:
URL: https://github.com/apache/tvm/pull/7123#issuecomment-756239725
Looking at the code, assuming you have thrust enabled, this should be
kernel0:
tqchen closed issue #7224:
URL: https://github.com/apache/tvm/issues/7224
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
tqchen commented on issue #7224:
URL: https://github.com/apache/tvm/issues/7224#issuecomment-756223503
The Module has been moved to ir.Module. Please open new discussion thread on
https://discuss.tvm.apache.org/ for further related questions
tqchen merged pull request #7045:
URL: https://github.com/apache/tvm/pull/7045
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
tqchen closed issue #6691:
URL: https://github.com/apache/tvm/issues/6691
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
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 9815ae2 [Arith] Simplify cast (#7045)
9815ae2 is
tqchen commented on pull request #7045:
URL: https://github.com/apache/tvm/pull/7045#issuecomment-756221819
Thanks @hzfan @junrushao1994 ! This is now merged
This is an automated message from the Apache Git Service.
To
rkimball commented on a change in pull request #7153:
URL: https://github.com/apache/tvm/pull/7153#discussion_r553391128
##
File path: cmake/modules/Libbacktrace.cmake
##
@@ -0,0 +1,18 @@
+include(ExternalProject)
+
+ExternalProject_Add(project_libbacktrace
+ PREFIX
dandingol03 opened a new issue #7224:
URL: https://github.com/apache/tvm/issues/7224
tvm: 0.8
error reproduce: `mod = relay.Module({})`
This is an automated message from the Apache Git Service.
To respond to the message,
AzazelHD edited a comment on issue #7217:
URL: https://github.com/apache/tvm/issues/7217#issuecomment-756106197
Yeah. I got a model that is on pytorch format. I used ``pytorch2keras``
library (python), and posted this error there. They said it must be an
``onnx2keras`` error.
The
AzazelHD commented on issue #7217:
URL: https://github.com/apache/tvm/issues/7217#issuecomment-756106197
Yeah. I got a model that is on pytorch format. I used ``pytorch2keras``
library (python), and posted this error there. They said it must be an
``onnx2keras`` error.
The model
TriLoo opened a new issue #7223:
URL: https://github.com/apache/tvm/issues/7223
I set `USE_MKL` to my mkl installed path in the `config.cmake` , however, I
got below errors:
```
error: ‘CBLAS_OFFSET’ does not name a type
...
error: ‘MKLStringToOffset’ was not declared in this
cailun01 opened a new pull request #7222:
URL: https://github.com/apache/tvm/pull/7222
Add tensorflow invert_permutation, which is used in MaskRCNN.
https://www.tensorflow.org/api_docs/python/tf/math/invert_permutation
64 matches
Mail list logo