limenghao opened a new pull request #7021:
URL: https://github.com/apache/tvm/pull/7021
Add features in the paper (AdaTune: Adaptive Tensor Program Compilation Made
Efficient).
Main changes:
1. In `./python/tvm/autotvm/measure/measure_methods.py`. Add an optional
argument
FrozenGene commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534805176
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
merrymercy commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534760628
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
merrymercy commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534760628
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
merrymercy commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534760628
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
merrymercy commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534755166
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
FrozenGene commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534731618
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
FrozenGene commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534731618
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
FrozenGene commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534724071
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
merrymercy commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534721692
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
FrozenGene commented on a change in pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#discussion_r534719806
##
File path: include/tvm/auto_scheduler/search_task.h
##
@@ -44,17 +44,16 @@ class HardwareParamsNode : public Object {
int cache_line_bytes;
//
merrymercy commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534702348
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
merrymercy commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534700837
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
merrymercy commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534699235
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
echuraev edited a comment on pull request #7013:
URL: https://github.com/apache/tvm/pull/7013#issuecomment-737682281
I've updated the PR. Now we will prefer IPv4 type of IP protocol between
IPv4 and IPv6. @tqchen please, take a look on it.
echuraev commented on pull request #7013:
URL: https://github.com/apache/tvm/pull/7013#issuecomment-737682281
I've updated the PR. Now we will prefer IPv4 type of IP protocol between
IPv4 and IPv6.
This is an automated
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534654075
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534653594
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534652925
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534652254
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534651971
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
tmoreau89 commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534651701
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,307 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
anijain2305 merged pull request #7016:
URL: https://github.com/apache/tvm/pull/7016
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
anijain2305 commented on pull request #7016:
URL: https://github.com/apache/tvm/pull/7016#issuecomment-737649557
Thanks @codeislife99 @trevor-m This is merged
This is an automated message from the Apache Git Service.
To
This is an automated email from the ASF dual-hosted git repository.
anijain2305 pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from a7bf979 [AutoScheduler] Support layout rewrite for whole networks
(#6987)
add 22a0877 Fix trt Test (#7016)
mbrookhart commented on pull request #7018:
URL: https://github.com/apache/tvm/pull/7018#issuecomment-737647798
Out of curiosity, why no nvptx?
This is an automated message from the Apache Git Service.
To respond to the
Meteorix commented on issue #7008:
URL: https://github.com/apache/tvm/issues/7008#issuecomment-737631890
@t-vi Thanks, I have read your PR and discuss thread. But when I use
`relay.build`, it still calls c++ build pipeline then c++ type inference.
merrymercy commented on pull request #7020:
URL: https://github.com/apache/tvm/pull/7020#issuecomment-737630159
cc @jcf94
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
merrymercy opened a new pull request #7020:
URL: https://github.com/apache/tvm/pull/7020
- Expose all fields of hardware parameters to its constructor
- Make `LogEstimatedLatency` as the default callback
This is an
comaniac commented on a change in pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#discussion_r534622889
##
File path: tutorials/auto_scheduler/tune_network_x86.py
##
@@ -0,0 +1,305 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more
merrymercy commented on pull request #7019:
URL: https://github.com/apache/tvm/pull/7019#issuecomment-737623657
cc @tqchen @junrushao1994 @jcf94 @minminsun
This is an automated message from the Apache Git Service.
To
merrymercy opened a new pull request #7019:
URL: https://github.com/apache/tvm/pull/7019
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.
lmzheng 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 a7bf979 [AutoScheduler] Support layout rewrite for
merrymercy merged pull request #6987:
URL: https://github.com/apache/tvm/pull/6987
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.
tqchen pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from e212f96 [Backend][Verilator] regression tests (#7000)
add 399e703 [µTVM] Modify reference VMs to support new µTVM
tqchen merged pull request #7001:
URL: https://github.com/apache/tvm/pull/7001
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
kevinthesun commented on a change in pull request #7018:
URL: https://github.com/apache/tvm/pull/7018#discussion_r534584859
##
File path: include/tvm/topi/transform.h
##
@@ -598,17 +598,69 @@ inline te::Tensor dynamic_strided_slice(const te::Tensor&
x, const te::Tensor& b
*
mbrookhart commented on a change in pull request #7018:
URL: https://github.com/apache/tvm/pull/7018#discussion_r534571294
##
File path: include/tvm/topi/transform.h
##
@@ -598,17 +598,69 @@ inline te::Tensor dynamic_strided_slice(const te::Tensor&
x, const te::Tensor& b
*
mbrookhart commented on a change in pull request #7018:
URL: https://github.com/apache/tvm/pull/7018#discussion_r534568965
##
File path: include/tvm/topi/transform.h
##
@@ -598,17 +598,69 @@ inline te::Tensor dynamic_strided_slice(const te::Tensor&
x, const te::Tensor& b
*
kevinthesun commented on pull request #6978:
URL: https://github.com/apache/tvm/pull/6978#issuecomment-737571529
https://github.com/apache/tvm/pull/7018
Fix for dynamic strided slice in topi to enable topk.
This is an
kevinthesun opened a new pull request #7018:
URL: https://github.com/apache/tvm/pull/7018
This fix also works for gpu argwhere.
@zhiics @anijain2305 @mbrookhart @Laurawly
This is an automated message from the
altanh commented on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737555189
Relevant issue on onnxruntime GitHub:
https://github.com/microsoft/onnxruntime/issues/5369
This is an automated message
tkonolige edited a comment on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737547902
The error message is:
```
OMP: Error #15: Initializing libomp.dylib, but found libiomp5.dylib already
initialized.
OMP: Hint This means that multiple copies of the
tkonolige commented on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737547902
The error message is:
```
OMP: Error #15: Initializing libomp.dylib, but found libiomp5.dylib already
initialized.
OMP: Hint This means that multiple copies of the OpenMP
tkonolige opened a new issue #7017:
URL: https://github.com/apache/tvm/issues/7017
#7010 was hard to debug because pytest printed nothing once a segfault
occurred. Forcing pytest to not capture output (via the `-s` flag) revealed the
cause of the segfault. It would be nice to have pytest
altanh commented on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737546289
per discussion with @tkonolige, we're pretty sure the abort is being caused
by `libomp` conflicts between different 3rd party libraries (e.g. PyTorch and
ONNX).
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534523208
##
File path: include/tvm/relay/transform.h
##
@@ -315,6 +323,12 @@ TVM_DLL Pass CanonicalizeOps();
*/
TVM_DLL Pass AlterOpLayout();
+/*!
+ * \brief
trevor-m commented on a change in pull request #7016:
URL: https://github.com/apache/tvm/pull/7016#discussion_r534501808
##
File path: tests/python/contrib/test_tensorrt.py
##
@@ -1050,26 +1050,26 @@ def test_tensorrt_dynamic_batch():
batches_to_test = [1, 1, 0, 2, 3, 0,
trevor-m commented on a change in pull request #7016:
URL: https://github.com/apache/tvm/pull/7016#discussion_r534501808
##
File path: tests/python/contrib/test_tensorrt.py
##
@@ -1050,26 +1050,26 @@ def test_tensorrt_dynamic_batch():
batches_to_test = [1, 1, 0, 2, 3, 0,
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534500373
##
File path: include/tvm/topi/transform.h
##
@@ -1400,6 +1400,74 @@ inline Tensor layout_transform(const Tensor& src, const
std::string& src_layout,
altanh commented on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737491908
I can't reproduce this locally on the current main branch
This is an automated message from the Apache Git Service.
To
trevor-m commented on a change in pull request #7016:
URL: https://github.com/apache/tvm/pull/7016#discussion_r534475764
##
File path: tests/python/contrib/test_tensorrt.py
##
@@ -1058,7 +1058,7 @@ def test_tensorrt_dynamic_batch():
mod = tvm.IRModule()
codeislife99 opened a new pull request #7016:
URL: https://github.com/apache/tvm/pull/7016
Fix TRT Test
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
This is an automated email from the ASF dual-hosted git repository.
moreau 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 e212f96 [Backend][Verilator] regression tests (#7000)
tmoreau89 merged pull request #7000:
URL: https://github.com/apache/tvm/pull/7000
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
tkonolige commented on pull request #7006:
URL: https://github.com/apache/tvm/pull/7006#issuecomment-737465838
This is a flakey test we've been seeing recently. Push a new commit to force
the ci to rerun.
This is an
slyubomirsky commented on pull request #7006:
URL: https://github.com/apache/tvm/pull/7006#issuecomment-737464345
The test failure seems spurious (in an operator test file I didn't modify).
This is an automated message from
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534440217
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534440217
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534439734
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534439734
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534439463
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534439106
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534438706
##
File path: tests/python/relay/test_auto_scheduler_layout_rewrite.py
##
@@ -0,0 +1,121 @@
+# Licensed to the Apache Software Foundation (ASF) under one
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534437675
##
File path: python/tvm/auto_scheduler/compute_dag.py
##
@@ -162,6 +162,23 @@ def infer_bound_from_state(self, state):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534437675
##
File path: python/tvm/auto_scheduler/compute_dag.py
##
@@ -162,6 +162,23 @@ def infer_bound_from_state(self, state):
areusch commented on a change in pull request #7002:
URL: https://github.com/apache/tvm/pull/7002#discussion_r534351281
##
File path: python/tvm/runtime/module.py
##
@@ -268,6 +268,11 @@ def export_library(self, file_name, fcompile=None,
addons=None, **kwargs):
comaniac commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534367774
##
File path: python/tvm/auto_scheduler/relay_integration.py
##
@@ -247,9 +271,74 @@ def auto_schedule_topi(outs, has_complex_op):
tkonolige commented on pull request #6910:
URL: https://github.com/apache/tvm/pull/6910#issuecomment-737403878
@tqchen Looks like all tests are passing now. Any other reviewers you would
suggest?
This is an automated
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534362427
##
File path: include/tvm/relay/transform.h
##
@@ -106,6 +106,14 @@ TVM_DLL Pass FoldConstant();
*/
TVM_DLL Pass FuseOps(int fuse_opt_level = -1);
tqchen commented on issue #7010:
URL: https://github.com/apache/tvm/issues/7010#issuecomment-737379507
https://ci.tlcpack.ai/job/tvm/job/main/250/execution/node/233/log/ another
one related
This is an automated message from
manupa-arm commented on a change in pull request #7002:
URL: https://github.com/apache/tvm/pull/7002#discussion_r534348967
##
File path: src/relay/backend/contrib/dnnl/codegen.cc
##
@@ -413,7 +413,8 @@ class DNNLModuleCodegen : public CSourceModuleCodegenBase {
// Create
areusch commented on pull request #7015:
URL: https://github.com/apache/tvm/pull/7015#issuecomment-737374668
@tqchen @jroesch
This is an automated message from the Apache Git Service.
To respond to the message, please log
areusch opened a new pull request #7015:
URL: https://github.com/apache/tvm/pull/7015
A few minor fixes to the tutorial which could help new users.
This is an automated message from the Apache Git Service.
To respond to the
zhiics commented on a change in pull request #7002:
URL: https://github.com/apache/tvm/pull/7002#discussion_r534335252
##
File path: src/relay/backend/build_module.cc
##
@@ -493,18 +493,14 @@ class RelayBuildModule : public runtime::ModuleNode {
// If we cannot decide
manupa-arm commented on a change in pull request #6950:
URL: https://github.com/apache/tvm/pull/6950#discussion_r534333446
##
File path: include/tvm/ir/module.h
##
@@ -56,11 +56,14 @@ class IRModuleNode : public Object {
Map type_definitions;
/*! \brief The source map
manupa-arm commented on a change in pull request #6950:
URL: https://github.com/apache/tvm/pull/6950#discussion_r534332379
##
File path: python/tvm/micro/build.py
##
@@ -162,7 +158,12 @@ def build_static_runtime(
libs.append(compiler.library(lib_build_dir, lib_srcs,
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534331753
##
File path: python/tvm/topi/nn/conv2d.py
##
@@ -371,8 +379,29 @@ def conv2d_nhwc(Input, Filter, stride, padding, dilation,
out_dtype="float32"):
merrymercy commented on a change in pull request #6987:
URL: https://github.com/apache/tvm/pull/6987#discussion_r534331214
##
File path: tests/python/relay/test_auto_scheduler_layout_rewrite.py
##
@@ -0,0 +1,122 @@
+# Licensed to the Apache Software Foundation (ASF) under one
manupa-arm commented on pull request #7002:
URL: https://github.com/apache/tvm/pull/7002#issuecomment-737363661
@areusch @tqchen @mbaret @zhiics
Please take a look when you have a minute :).
This is an automated message
t-vi commented on issue #7008:
URL: https://github.com/apache/tvm/issues/7008#issuecomment-737327528
#6900 is my attempt at getting incremental type inference into the PyToch
frontend. It will need some serious cleanup, but I think it is useful. It
removes the N² thing of re-checking
hgt312 opened a new pull request #7014:
URL: https://github.com/apache/tvm/pull/7014
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 reviews from
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 636739a [µTVM] Minor fixes to the Reference VM
tqchen merged pull request #7012:
URL: https://github.com/apache/tvm/pull/7012
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 pull request #7013:
URL: https://github.com/apache/tvm/pull/7013#issuecomment-737244613
Thanks @echuraev It is indeed good to support IPv6, although the current
change could make rpc tracker unavailable from other machines. The second best
option is indeed to check and
tqchen commented on a change in pull request #7013:
URL: https://github.com/apache/tvm/pull/7013#discussion_r534182868
##
File path: python/tvm/exec/rpc_tracker.py
##
@@ -33,7 +33,7 @@ def main(args):
if __name__ == "__main__":
parser = argparse.ArgumentParser()
-
echuraev commented on pull request #7013:
URL: https://github.com/apache/tvm/pull/7013#issuecomment-737143035
And just a question to this PR. Maybe the better option is to set IPv4 type
for all rpc connections? Or sometimes it is reasonable to have IPv6?
echuraev opened a new pull request #7013:
URL: https://github.com/apache/tvm/pull/7013
This change fix problem with version of IP protocol on MacOS.
Previous the rpc_tracker and query_rpc_tracker were not able connect to
each other with default hostnames.
The root cause was in
88 matches
Mail list logo