junrushao1994 commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456194801
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -161,16 +202,116 @@ def fuse(self, stage, iters):
The Stage to be
lsy643 commented on pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076#issuecomment-659854233
@zhiics @kevinthesun I am not trying to support heterogeneous execution in
this PR. What I want is to make a model with operators all on gpu and dynamic
input can compile
junrushao1994 commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456230063
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -161,16 +202,116 @@ def fuse(self, stage, iters):
The Stage to be
tqchen commented on pull request #6081:
URL: https://github.com/apache/incubator-tvm/pull/6081#issuecomment-659794824
THanks @kparzysz-quic !
This is an automated message from the Apache Git Service.
To respond to the
This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from b85c239 Refine LSTMBlockCell to support dynamic rnn (#5963)
add 6e1b09e [TOPI] Fix the filter width
liangfu edited a comment on pull request #5842:
URL: https://github.com/apache/incubator-tvm/pull/5842#issuecomment-659829041
Thanks for the update @zhanghaohit , please rebase upon latest master branch
and resolve the conflicts.
liangfu commented on pull request #5842:
URL: https://github.com/apache/incubator-tvm/pull/5842#issuecomment-659829041
Please rebase upon latest master branch and resolve the conflicts.
This is an automated message from the
merrymercy commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456228059
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -161,16 +202,116 @@ def fuse(self, stage, iters):
The Stage to be
zhiics commented on pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076#issuecomment-659865400
The way you currently do is more like a hack to make the example work. We
need to more systematically consider different instructions/ir nodes.
ShapeFuncs are always on
zhiics closed pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076
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
liangfu commented on a change in pull request #9:
URL: https://github.com/apache/incubator-tvm-vta/pull/9#discussion_r456229024
##
File path: hardware/intelfocl/src/vta.cl
##
@@ -0,0 +1,341 @@
+#pragma OPENCL EXTENSION cl_intel_channels: enable
Review comment:
Just
FrozenGene commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456188551
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -161,16 +202,116 @@ def fuse(self, stage, iters):
The Stage to be
csullivan opened a new pull request #6085:
URL: https://github.com/apache/incubator-tvm/pull/6085
Add support for `pad_to_max_output_size` field in NMSv4 from Tensorflow
which results in two outputs:
1. Indices tensor (vector of length max_num_indices.
2. Scalar tensor indicating
tqchen merged pull request #6081:
URL: https://github.com/apache/incubator-tvm/pull/6081
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
FrozenGene commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r456186767
##
File path: tests/python/contrib/test_arm_compute_lib/infrastructure.py
##
@@ -0,0 +1,197 @@
+# Licensed to the Apache Software Foundation
merrymercy commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456228059
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -161,16 +202,116 @@ def fuse(self, stage, iters):
The Stage to be
junrushao1994 commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456206595
##
File path: tests/python/unittest/test_auto_scheduler_loop_state.py
##
@@ -61,5 +61,79 @@ def test_split_fuse_reorder():
assert
jcf94 opened a new pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
FrozenGene opened a new pull request #6072:
URL: https://github.com/apache/incubator-tvm/pull/6072
We have unnecessary std::cout appeared in
https://github.com/apache/incubator-tvm/pull/6047
@mbrookhart
This is
lixiaoquan commented on pull request #5963:
URL: https://github.com/apache/incubator-tvm/pull/5963#issuecomment-659206874
ping @zhiics Could you help to review and merge? Thanks
This is an automated message from the Apache
This is an automated email from the ASF dual-hosted git repository.
liangfu pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from dff715a Fix error message in Buffer::vstore, NFC (#6056)
add 3fda73f [RUNTIME][CRT] init
liangfu merged pull request #6044:
URL: https://github.com/apache/incubator-tvm/pull/6044
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
poyenc opened a new pull request #6075:
URL: https://github.com/apache/incubator-tvm/pull/6075
This pull request add `PassEnabled()` calls for the dependency passes to
avoid execution of any disable passes.
@tqchen, @jroesch please help review, thanks!
lsy643 opened a new pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076
Currently, when we want to compile and execute a network with dynamic input,
only using `llvm` target can work as expected.
If the target is as `CUDA`, the compilation part works fine,
lhutton1 commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r455643707
##
File path: docs/deploy/arm_compute_lib.rst
##
@@ -0,0 +1,135 @@
+.. Licensed to the Apache Software Foundation (ASF) under one
+or more
lhutton1 commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r455643955
##
File path: docs/deploy/arm_compute_lib.rst
##
@@ -0,0 +1,135 @@
+.. Licensed to the Apache Software Foundation (ASF) under one
+or more
wrongtest edited a comment on pull request #6062:
URL: https://github.com/apache/incubator-tvm/pull/6062#issuecomment-659137025
> Thanks for the PR! I have two questions other than the comments:
>
> 1. How to trigger this change (i.e., `to_batch=false`) for an end user? It
seems to
liangfu commented on pull request #6044:
URL: https://github.com/apache/incubator-tvm/pull/6044#issuecomment-659329743
Thanks @windclarion for contributing. Thanks @tqchen for reviewing. This is
now merged.
This is an
poyenc opened a new issue #6074:
URL: https://github.com/apache/incubator-tvm/issues/6074
## Issue
In
[`SequentialNode::operator()`](https://github.com/apache/incubator-tvm/blob/8578096853eec5711bfcc9a3a68145fd12a135cb/src/ir/transform.cc#L364),
before execute each passes, we do not
jcf94 opened a new pull request #6077:
URL: https://github.com/apache/incubator-tvm/pull/6077
For full upstream plan, see [Ansor
RFC](https://discuss.tvm.ai/t/rfc-ansor-an-auto-scheduler-for-tvm-autotvm-v2-0/7005/32).
In this PR, we bring the RPCRunner. This uses RPC call to
jwfromm commented on a change in pull request #5235:
URL: https://github.com/apache/incubator-tvm/pull/5235#discussion_r455845167
##
File path: src/te/schedule/operation_inline.cc
##
@@ -63,7 +63,9 @@ class OperationInliner final : public StmtExprMutator {
} else {
yzhliu opened a new pull request #6078:
URL: https://github.com/apache/incubator-tvm/pull/6078
Co-authored-by: Sergei Grechanik
This is the PR that aim to remove the intermediate large Jacobian tensor in
te.autodiff, as well as to do some optimizations such as inline. We have made
FrozenGene commented on pull request #6077:
URL: https://github.com/apache/incubator-tvm/pull/6077#issuecomment-659512977
@jcf94 Gental reminder: To reproduce Ansor paper's result, we should turn on
clflush (for x86 cpu) and non empty array. clflush has been merged, you should
consider to
comaniac commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r455863170
##
File path: python/tvm/relay/op/contrib/arm_compute_lib.py
##
@@ -0,0 +1,119 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+#
eric-haibin-lin commented on issue #6015:
URL: https://github.com/apache/incubator-tvm/issues/6015#issuecomment-659551366
I'll take a stab
This is an automated message from the Apache Git Service.
To respond to the message,
zhiics merged pull request #6072:
URL: https://github.com/apache/incubator-tvm/pull/6072
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
zhiics commented on pull request #6072:
URL: https://github.com/apache/incubator-tvm/pull/6072#issuecomment-659559802
Thanks @FrozenGene @junrushao1994
This is an automated message from the Apache Git Service.
To respond to
This is an automated email from the ASF dual-hosted git repository.
zhic pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 3fda73f [RUNTIME][CRT] init TVMPackedFunc's name (#6044)
add 5b46192 Remove unnecessary std::cout
mbrookhart commented on pull request #6072:
URL: https://github.com/apache/incubator-tvm/pull/6072#issuecomment-659563554
Oops! Thanks for catching it.
This is an automated message from the Apache Git Service.
To respond to
lhutton1 commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r455957536
##
File path: tests/python/contrib/test_arm_compute_lib/test_runtime.py
##
@@ -0,0 +1,98 @@
+# Licensed to the Apache Software Foundation (ASF)
junrushao1994 commented on pull request #6077:
URL: https://github.com/apache/incubator-tvm/pull/6077#issuecomment-659624761
Thanks for the PR! Will take a look this weekend
This is an automated message from the Apache Git
kevinthesun commented on pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076#issuecomment-659575125
@zhiics
This is an automated message from the Apache Git Service.
To respond to the message, please
eric-haibin-lin opened a new pull request #6079:
URL: https://github.com/apache/incubator-tvm/pull/6079
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
lhutton1 commented on pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#issuecomment-659622884
Not too sure what happened to the Mac build, looks unrelated.
This is an automated message from the
huajsj commented on pull request #6070:
URL: https://github.com/apache/incubator-tvm/pull/6070#issuecomment-659610749
@tmoreau89 @liangfu can you help to do a review for this patch? thanks.
Regards
Hua
This is an
kevinthesun merged pull request #6048:
URL: https://github.com/apache/incubator-tvm/pull/6048
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.
kevinthesun pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 5b46192 Remove unnecessary std::cout (#6072)
add 9f6544b [AutoTVM][BugFix] Fix variable name
kevinthesun commented on pull request #6048:
URL: https://github.com/apache/incubator-tvm/pull/6048#issuecomment-659577789
Thanks @ymwangg @comaniac
This is an automated message from the Apache Git Service.
To respond to
zhiics commented on pull request #6076:
URL: https://github.com/apache/incubator-tvm/pull/6076#issuecomment-659578197
Let's not do this because it cannot solve the problem. Dynamic shape support
on GPU needs heterogeneous execution. We have worked on it and will refine the
code and
mbrookhart opened a new pull request #6080:
URL: https://github.com/apache/incubator-tvm/pull/6080
cc @icemelon9 @MarisaKirisame @electriclilies
This is an automated message from the Apache Git Service.
To respond to the
tmoreau89 merged pull request #6070:
URL: https://github.com/apache/incubator-tvm/pull/6070
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
This is an automated email from the ASF dual-hosted git repository.
moreau pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 9f6544b [AutoTVM][BugFix] Fix variable name conflict with OpenCL
keyword (#6048)
add 910ad00 [VTA]
comaniac commented on pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#issuecomment-659660157
> Not too sure what happened to the Mac build, looks unrelated.
Should not be an issue. You could use `git commit --allow-empty` to
re-trigger the CI.
tqchen commented on pull request #6075:
URL: https://github.com/apache/incubator-tvm/pull/6075#issuecomment-659659462
Thanks for the discussion. In this particular case, I think the behavior is
expected. Because the dependency itself might be required in order to run the
pass
poyenc closed pull request #6075:
URL: https://github.com/apache/incubator-tvm/pull/6075
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
poyenc closed issue #6074:
URL: https://github.com/apache/incubator-tvm/issues/6074
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
junrushao1994 commented on pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#issuecomment-659627481
Just noticed that #6077 is also titled as "part 1", shall we make one of
them part 2 instead?
This
kparzysz-quic opened a new pull request #6081:
URL: https://github.com/apache/incubator-tvm/pull/6081
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
tqchen merged pull request #6043:
URL: https://github.com/apache/incubator-tvm/pull/6043
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
This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 910ad00 [VTA] Fix FSIM Compile Error. (#6070)
add 30415e5 [ARITH] Improve vector simplification for
tqchen commented on issue #6074:
URL: https://github.com/apache/incubator-tvm/issues/6074#issuecomment-659650358
In this particular case, I think the behavior is expected. Because the
dependency itself might be required in order to run the pass
tqchen edited a comment on issue #6074:
URL: https://github.com/apache/incubator-tvm/issues/6074#issuecomment-659650358
Thanks for the discussion. In this particular case, I think the behavior is
expected. Because the dependency itself might be required in order to run the
pass
zhiics commented on a change in pull request #5915:
URL: https://github.com/apache/incubator-tvm/pull/5915#discussion_r456060646
##
File path: src/relay/backend/contrib/arm_compute_lib/codegen_acl.h
##
@@ -0,0 +1,107 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF)
eric-haibin-lin closed pull request #6079:
URL: https://github.com/apache/incubator-tvm/pull/6079
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
tqchen commented on a change in pull request #5235:
URL: https://github.com/apache/incubator-tvm/pull/5235#discussion_r456060147
##
File path: tests/python/relay/test_pass_fuse_ops.py
##
@@ -621,6 +621,81 @@ def expected():
after = run_opt_pass(expected(),
anijain2305 commented on a change in pull request #5980:
URL: https://github.com/apache/incubator-tvm/pull/5980#discussion_r456125120
##
File path: python/tvm/tir/op.py
##
@@ -965,6 +965,34 @@ def popcount(x):
"""
return call_intrin(x.dtype, "tir.popcount", x)
+def
anijain2305 commented on a change in pull request #5980:
URL: https://github.com/apache/incubator-tvm/pull/5980#discussion_r456134760
##
File path: include/tvm/tir/builtin.h
##
@@ -92,6 +92,14 @@ TVM_DLL const Op& shift_right();
*/
TVM_DLL const Op& large_uint_imm();
+/*!
anijain2305 commented on pull request #5980:
URL: https://github.com/apache/incubator-tvm/pull/5980#issuecomment-659734746
@tmoreau89 @kparzysz-quic Please review again and approve explicitly.
This is an automated message
tmoreau89 commented on a change in pull request #5980:
URL: https://github.com/apache/incubator-tvm/pull/5980#discussion_r456142538
##
File path: include/tvm/tir/builtin.h
##
@@ -92,6 +92,14 @@ TVM_DLL const Op& shift_right();
*/
TVM_DLL const Op& large_uint_imm();
+/*!
+
tkonolige opened a new pull request #6083:
URL: https://github.com/apache/incubator-tvm/pull/6083
Python 3.8 changes the default way multiprocessing creates new processes on
macOS from forking to spawing. Spawning requires all objects to be picklable.
Nested functions and lambdas are not
tkonolige commented on pull request #6083:
URL: https://github.com/apache/incubator-tvm/pull/6083#issuecomment-659743769
@tqchen @junrushao1994
This is an automated message from the Apache Git Service.
To respond to the
tmoreau89 commented on pull request #5980:
URL: https://github.com/apache/incubator-tvm/pull/5980#issuecomment-659743763
@giuseros thanks for adding the test case, and thank you @anijain2305 for
the thorough review. I suggest to follow through with the naming tweak and this
PR should be
icemelon9 opened a new pull request #6082:
URL: https://github.com/apache/incubator-tvm/pull/6082
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
kevinthesun merged pull request #5963:
URL: https://github.com/apache/incubator-tvm/pull/5963
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
kevinthesun commented on pull request #5963:
URL: https://github.com/apache/incubator-tvm/pull/5963#issuecomment-659707295
Thanks @lixiaoquan @yongwww
This is an automated message from the Apache Git Service.
To respond to
This is an automated email from the ASF dual-hosted git repository.
kevinthesun pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 30415e5 [ARITH] Improve vector simplification for float operands
(#6043)
add b85c239 Refine
abergeron commented on pull request #6065:
URL: https://github.com/apache/incubator-tvm/pull/6065#issuecomment-659742964
It seems the master is broken, because some rust test is failing and I
didn't touch any of the rust code.
It also seems that the macOS build is flaky because it
comaniac commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456141719
##
File path: src/auto_scheduler/loop_state.cc
##
@@ -90,12 +90,69 @@ Stage::Stage(te::Operation op, StageKind op_type, const
Array& iters,
comaniac commented on pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#issuecomment-659756713
> Just noticed that #6077 is also titled as "part 1", shall we make one of
them part 2 instead?
Part 1 will have several PRs. "Phase 1" might be a better name to
comaniac edited a comment on pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#issuecomment-659756713
> Just noticed that #6077 is also titled as "part 1", shall we make one of
them part 2 instead?
Part 1 will have several PRs. "Phase 1" might be a better
leonwanghui commented on pull request #5892:
URL: https://github.com/apache/incubator-tvm/pull/5892#issuecomment-659758783
cc @tqchen
This is an automated message from the Apache Git Service.
To respond to the message,
anijain2305 opened a new pull request #6084:
URL: https://github.com/apache/incubator-tvm/pull/6084
@tqchen
Issue - https://github.com/apache/incubator-tvm/issues/6064
The error earlier was for single element out of 3456 elements. I think this
is because of rounding. For now,
jcf94 commented on pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#issuecomment-659776003
> > Just noticed that #6077 is also titled as "part 1", shall we make one of
them part 2 instead?
>
> Part 1 will have several PRs. "Phase 1" might be a better word
jcf94 commented on pull request #6077:
URL: https://github.com/apache/incubator-tvm/pull/6077#issuecomment-659778310
> @jcf94 Gental reminder: To reproduce Ansor paper's result, we should turn
on clflush (for x86 cpu) and non empty array. clflush has been merged, you
should consider to
jcf94 commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456174561
##
File path: src/auto_scheduler/measure_record.cc
##
@@ -169,6 +206,18 @@ struct Handler<::tvm::Array<::tvm::auto_scheduler::Step>> {
zhanghaohit commented on pull request #5842:
URL: https://github.com/apache/incubator-tvm/pull/5842#issuecomment-659790884
> Note that I've also uploaded the Pynq bitstream given the ISA changes
here: https://github.com/uwsampl/vta-distro/tree/master/bitstreams/pynq/0.0.2
>
> It
zhanghaohit commented on pull request #5842:
URL: https://github.com/apache/incubator-tvm/pull/5842#issuecomment-659791729
@tmoreau89 @vegaluisjose @huajsj @pasqoc
Any comments are welcome. Thanks.
This is an automated
zhanghaohit commented on pull request #9:
URL: https://github.com/apache/incubator-tvm-vta/pull/9#issuecomment-659792207
@tmoreau89 @liangfu any other comments? Thanks.
This is an automated message from the Apache Git
jcf94 commented on a change in pull request #6073:
URL: https://github.com/apache/incubator-tvm/pull/6073#discussion_r456181569
##
File path: tests/python/unittest/test_auto_scheduler_loop_state.py
##
@@ -61,5 +61,79 @@ def test_split_fuse_reorder():
assert
89 matches
Mail list logo