comaniac commented on pull request #9875:
URL: https://github.com/apache/tvm/pull/9875#issuecomment-1007904926
Thanks for the RFC. Would you please file a formal RFC here:
https://github.com/apache/tvm-rfcs along with the upstream plan? For example,
you have some TODOs in this PR, so it wo
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780642764
##
File path: src/tir/schedule/analysis/analysis.cc
##
@@ -1376,5 +1376,333 @@ void CheckStorageScope(const ScheduleState& self,
String storage_scope)
junrushao1994 commented on pull request #9875:
URL: https://github.com/apache/tvm/pull/9875#issuecomment-1007900534
CC: @zxybazh @comaniac @merrymercy
--
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.
junrushao 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 ab8a106 [MetaSchedule] Add Per-Store-Feature (#9860
junrushao1994 merged pull request #9860:
URL: https://github.com/apache/tvm/pull/9860
--
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-unsu
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780639226
##
File path: src/tir/schedule/analysis.h
##
@@ -442,6 +442,79 @@ bool CanComputeAt(const ScheduleState& self, const
StmtSRef& block_sref, const S
boo
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780639158
##
File path: src/tir/schedule/transform.cc
##
@@ -136,5 +136,52 @@ void LeafBlockRemovalPlan(const ScheduleState& self, const
StmtSRef& leaf_block_
Hzfengsy commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780634559
##
File path: include/tvm/tir/function.h
##
@@ -187,6 +187,56 @@ class LinkedParam : public ObjectRef {
TVM_DEFINE_OBJECT_REF_COW_METHOD(LinkedParamNode);
mshr-h opened a new pull request #9879:
URL: https://github.com/apache/tvm/pull/9879
because we don't need it for CI.
--
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 un
mshr-h opened a new pull request #9878:
URL: https://github.com/apache/tvm/pull/9878
Purpose of this change is to remove rust docs which we don't need for CI.
`/opt/rust` will be reduced from 1.0GB->551MB.
[Profiles - The rustup
book](https://rust-lang.github.io/rustup/concepts/pr
mshr-h closed pull request #9877:
URL: https://github.com/apache/tvm/pull/9877
--
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...
mshr-h opened a new pull request #9877:
URL: https://github.com/apache/tvm/pull/9877
Purpose of this change is to remove rust docs which we don't need for CI.
`/opt/rust` will be reduced from 1.0GB->551MB.
[Profiles - The rustup
book](https://rust-lang.github.io/rustup/concepts/pr
huajsj opened a new pull request #9876:
URL: https://github.com/apache/tvm/pull/9876
[Issue]
When the input shape not compatible with reshaped shape and not dynamic
in topi reshape, the two shape sum value will get printed out as an error
message, but such shape sum value is not hel
noobotdj opened a new pull request #9875:
URL: https://github.com/apache/tvm/pull/9875
We write an
Pre-RFC(https://discuss.tvm.apache.org/t/rfc-familyseer-a-new-search-method-for-auto-scheduler/11877
) and waiting for comments from reviewers.
--
This is an automated message from the
Raghav-Chakravarthy commented on pull request #9770:
URL: https://github.com/apache/tvm/pull/9770#issuecomment-1007870234
@huajsj
Hello,
I have fixed the build issue. Please help me merge the commit.
Thanks,
Raghav
--
This is an automated message from the Apache Git S
jinhongyii commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780620536
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
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 f6f252f [TOPI] Support grouped conv1d (#9832)
add 38f0239 [CI] Update to PyTorch v1.10 in GPU image (#9866)
No new
masahi commented on issue #9349:
URL: https://github.com/apache/tvm/issues/9349#issuecomment-1007864029
Upgrade to PT v1.10 is now complete.
--
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
masahi merged pull request #9866:
URL: https://github.com/apache/tvm/pull/9866
--
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...
masahi closed issue #9349:
URL: https://github.com/apache/tvm/issues/9349
--
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.
masahi closed issue #9713:
URL: https://github.com/apache/tvm/issues/9713
--
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.
electriclilies commented on pull request #9826:
URL: https://github.com/apache/tvm/pull/9826#issuecomment-1007862755
Rebased with #9874
--
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
electriclilies opened a new pull request #9874:
URL: https://github.com/apache/tvm/pull/9874
In this PR, I add some JSON converters. The motivation for this change is
#9826-- I need to visit the virtual_device_ field in the AttrVisitor of many
Relay nodes. Unfortunately, this breaks import
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780605458
##
File path: src/tir/schedule/analysis.h
##
@@ -442,6 +442,79 @@ bool CanComputeAt(const ScheduleState& self, const
StmtSRef& block_sref, const S
boo
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780605351
##
File path: src/tir/schedule/analysis.h
##
@@ -442,6 +442,79 @@ bool CanComputeAt(const ScheduleState& self, const
StmtSRef& block_sref, const S
boo
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780603282
##
File path: src/tir/schedule/analysis.h
##
@@ -442,6 +442,79 @@ bool CanComputeAt(const ScheduleState& self, const
StmtSRef& block_sref, const S
boo
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780598498
##
File path: src/tir/schedule/state.cc
##
@@ -201,9 +201,8 @@ class BlockInfoCollector : private StmtVisitor {
bool is_root_block = srefs_.empty()
vinx13 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780598226
##
File path: src/tir/schedule/state.cc
##
@@ -201,9 +201,8 @@ class BlockInfoCollector : private StmtVisitor {
bool is_root_block = srefs_.empty();
junrushao1994 commented on a change in pull request #9866:
URL: https://github.com/apache/tvm/pull/9866#discussion_r780594595
##
File path: tests/scripts/task_config_build_gpu.sh
##
@@ -30,7 +30,7 @@ echo set\(USE_VULKAN ON\) >> config.cmake
echo set\(USE_OPENGL ON\) >> config
junrushao1994 commented on a change in pull request #9866:
URL: https://github.com/apache/tvm/pull/9866#discussion_r780594405
##
File path: tests/scripts/task_config_build_gpu.sh
##
@@ -30,7 +30,7 @@ echo set\(USE_VULKAN ON\) >> config.cmake
echo set\(USE_OPENGL ON\) >> config
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780592950
##
File path: src/tir/schedule/state.cc
##
@@ -201,9 +201,8 @@ class BlockInfoCollector : private StmtVisitor {
bool is_root_block = srefs_.empty()
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780592805
##
File path: src/tir/schedule/schedule.cc
##
@@ -183,6 +183,20 @@ TVM_REGISTER_GLOBAL("tir.schedule.ScheduleStorageAlign")
TVM_REGISTER_GLOBAL("tir.sc
tkonolige commented on pull request #9872:
URL: https://github.com/apache/tvm/pull/9872#issuecomment-1007828306
So things are a little more complicated here than it appears at first
glance. And error thrown by tvm via the LOG and CHECK infrastructure constructs
a TVMError with a backtrace.
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780590829
##
File path: python/tvm/tir/schedule/schedule.py
##
@@ -1733,6 +1733,258 @@ def after_set_scope(
## Schedule: Blockize & Tensorize #
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589809
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589725
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589634
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589562
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589481
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780589224
##
File path: src/tir/ir/function.cc
##
@@ -64,6 +64,67 @@ FuncType PrimFuncNode::func_type_annotation() const {
TVM_REGISTER_NODE_TYPE(PrimFuncNode)
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780588745
##
File path: python/tvm/tir/schedule/schedule.py
##
@@ -1733,6 +1733,258 @@ def after_set_scope(
## Schedule: Blockize & Tensorize #
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780588169
##
File path: include/tvm/tir/schedule/schedule.h
##
@@ -465,6 +465,25 @@ class ScheduleNode : public runtime::Object {
*/
virtual void SetScope(
driazati commented on pull request #9872:
URL: https://github.com/apache/tvm/pull/9872#issuecomment-1007823006
> The colors look great! How does this interact with backtraces in python?
Are only the c++ parts of the backtrace colored?
It doesn't touch Python at all since those are (I
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780587954
##
File path: include/tvm/tir/function.h
##
@@ -187,6 +187,56 @@ class LinkedParam : public ObjectRef {
TVM_DEFINE_OBJECT_REF_COW_METHOD(LinkedParamN
junrushao1994 commented on a change in pull request #9871:
URL: https://github.com/apache/tvm/pull/9871#discussion_r780587280
##
File path: include/tvm/tir/function.h
##
@@ -187,6 +187,56 @@ class LinkedParam : public ObjectRef {
TVM_DEFINE_OBJECT_REF_COW_METHOD(LinkedParamN
junrushao1994 commented on pull request #9872:
URL: https://github.com/apache/tvm/pull/9872#issuecomment-1007820996
BTW, there is an pre-RFC discussing logging:
https://discuss.tvm.apache.org/t/rfc-better-tvm-logger-on-both-c-and-python-side/11734/13.
Let's post this PR to the thread for m
tkonolige opened a new pull request #9873:
URL: https://github.com/apache/tvm/pull/9873
Change conv3d to use generic conv implementation which supports grouped
convolutions. Also, remove support for non-float16 tensorcore operations as
they cause large degradation in accuracy. Generic conv
junrushao1994 commented on pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#issuecomment-1007818731
@jinhongyii @comaniac Comments addressed! Please take another look :-)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on
This is an automated email from the ASF dual-hosted git repository.
mbrookhart pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from afc29e6 [microNPU][2b] Create CascaderGraphs from TE graphs (#9471)
add f6f252f [TOPI] Support grouped conv1d (
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780583766
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
mbrookhart commented on pull request #9832:
URL: https://github.com/apache/tvm/pull/9832#issuecomment-1007818008
Thanks, @tkonolige !
--
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
mbrookhart merged pull request #9832:
URL: https://github.com/apache/tvm/pull/9832
--
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-unsubsc
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780583446
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
tkonolige commented on pull request #9872:
URL: https://github.com/apache/tvm/pull/9872#issuecomment-1007817625
The colors look great! How does this interact with backtraces in python? Are
only the c++ parts of the backtrace colored?
--
This is an automated message from the Apache Git Se
junrushao1994 commented on pull request #9872:
URL: https://github.com/apache/tvm/pull/9872#issuecomment-1007796282
adding colors is certainly super color feature! BTW, is it possible not to
introduce platform dependent code? Note that we have rang as the 3rdparty repo
--
This is an auto
driazati opened a new pull request #9872:
URL: https://github.com/apache/tvm/pull/9872
This adds colors (which match GDB's output) to files / symbols when printing
exceptions from C++ if stdout is a terminal. This also adds an environment
variable `TVM_BACKTRACE_LIMIT` which can be used to
vinx13 opened a new pull request #9871:
URL: https://github.com/apache/tvm/pull/9871
This PR added two schedule primitives, blockize, and tensorize to TensorIR.
cc @junrushao1994 @spectrometerHBH @Hzfengsy @masahi @comaniac @tqchen
@csullivan
--
This is an automated message from
michalpiszczek commented on pull request #9870:
URL: https://github.com/apache/tvm/pull/9870#issuecomment-1007781704
@mbs-octoml @mbrookhart PTAL
--
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
AndrewZhaoLuo opened a new pull request #9869:
URL: https://github.com/apache/tvm/pull/9869
DynamicToStatic has the same issue as found in #9735 as it incrementally
tries to make dynamic things static throughout the graph, running type
inference incrementally along the way. DynamicToStatic
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780546700
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
tkonolige opened a new issue #9868:
URL: https://github.com/apache/tvm/issues/9868
```
$ ./tests/scripts/task_python_integration.sh
... output removed ...
=== test session
starts ===
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780542125
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780541080
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780540511
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
tkonolige opened a new pull request #9867:
URL: https://github.com/apache/tvm/pull/9867
@areusch
--
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: c
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780535499
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
comaniac commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780528430
##
File path:
tests/python/unittest/test_meta_schedule_feature_extractor_per_store_feature.py
##
@@ -0,0 +1,1554 @@
+# Licensed to the Apache Software Found
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780523605
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780523294
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780523071
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
junrushao1994 commented on a change in pull request #9860:
URL: https://github.com/apache/tvm/pull/9860#discussion_r780522736
##
File path: src/meta_schedule/feature_extractor/per_store_feature.cc
##
@@ -0,0 +1,1307 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) un
tqchen commented on pull request #9842:
URL: https://github.com/apache/tvm/pull/9842#issuecomment-1007720436
TVM_LOG_DEBUG might be a good candidate given it is used to control DCHECK
etc. I agree that iterator concurrent update is a problem of stl containers as
well so we should not inclu
electriclilies commented on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007709472
Alright, I'll do the updater and open a new PR. Closing this now
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to Gi
electriclilies closed pull request #9861:
URL: https://github.com/apache/tvm/pull/9861
--
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-uns
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed one
solution to the particular problem as well.
One would need to think about the implication th
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
tqchen commented on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though, since
tqchen edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007698755
Patch up specialized handling for a particular object and key is indeed a
solution to the current problem as well.
One would need to think about the implication though,
driazati commented on a change in pull request #49:
URL: https://github.com/apache/tvm-rfcs/pull/49#discussion_r780471477
##
File path: rfcs/0049-managed-jenkins-infrastructure-for-tvm.md
##
@@ -0,0 +1,136 @@
+# Managed Jenkins Infrastructure for TVM
+
+- Feature Name: `managed
Lunderberg commented on pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#issuecomment-1007651198
Rebasing onto main to resolve merge conflicts. Apologies in advance for any
in-progress reviews that get orphaned by the rebase.
--
This is an automated message from the Apac
electriclilies edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007648766
Another possible solution is adding a HasValue method and a list of
defaults.
```
Map defualt_values = ["virtual_device_", "0"]
bool HasValue(const c
electriclilies commented on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007648766
Another possible solution is adding a HasValue method and a list of
defaults.
```
Map defualt_values = ["virtual_device_", "0"]
bool HasValue(const char* ke
electriclilies edited a comment on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007624056
OK, thanks for the thoughts.
What do you think about just letting the virtual_device_ have a default? So,
I'd add something like this:
```
if key == "
electriclilies commented on pull request #9861:
URL: https://github.com/apache/tvm/pull/9861#issuecomment-1007624056
OK, thanks for the thoughts.
What do you think about just letting the virtual_device_ have a default? So,
I'd add something like this:
```
if key == "virtual
areusch commented on a change in pull request #49:
URL: https://github.com/apache/tvm-rfcs/pull/49#discussion_r780372722
##
File path: rfcs/0049-managed-jenkins-infrastructure-for-tvm.md
##
@@ -0,0 +1,136 @@
+# Managed Jenkins Infrastructure for TVM
+
+- Feature Name: `managed_
areusch commented on a change in pull request #49:
URL: https://github.com/apache/tvm-rfcs/pull/49#discussion_r780368372
##
File path: rfcs/0049-managed-jenkins-infrastructure-for-tvm.md
##
@@ -0,0 +1,136 @@
+# Managed Jenkins Infrastructure for TVM
+
+- Feature Name: `managed_
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 47d6eb2 [LLVM] LLVM codegen debug utilities (#9857)
add afc29e6 [microNPU][2b] Create CascaderGraphs from TE graph
areusch commented on a change in pull request #49:
URL: https://github.com/apache/tvm-rfcs/pull/49#discussion_r780363270
##
File path: rfcs/0049-managed-jenkins-infrastructure-for-tvm.md
##
@@ -0,0 +1,136 @@
+# Managed Jenkins Infrastructure for TVM
+
+- Feature Name: `managed_
Mousius merged pull request #9471:
URL: https://github.com/apache/tvm/pull/9471
--
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..
areusch commented on a change in pull request #49:
URL: https://github.com/apache/tvm-rfcs/pull/49#discussion_r780359577
##
File path: rfcs/0049-managed-jenkins-infrastructure-for-tvm.md
##
@@ -0,0 +1,136 @@
+# Managed Jenkins Infrastructure for TVM
+
+- Feature Name: `managed_
Lunderberg commented on a change in pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#discussion_r780353202
##
File path: include/tvm/tir/buffer.h
##
@@ -55,8 +55,48 @@ class BufferNode : public Object {
Var data;
/*! \brief data type in the content of the
areusch commented on pull request #50:
URL: https://github.com/apache/tvm-rfcs/pull/50#issuecomment-1007513046
@denise-k @tqchen @jroesch @junrushao1994 @hogepodge @masahi @vinx13
@comaniac @yzhliu @zhiics @icemelon @Hzfengsy @kparzysz-quic @mdw-octoml
@mousius @u99127 @manupa-arm @leandro
Mousius commented on a change in pull request #9471:
URL: https://github.com/apache/tvm/pull/9471#discussion_r780336416
##
File path: python/tvm/relay/backend/contrib/ethosu/te/convolution.py
##
@@ -164,5 +168,126 @@ def conv2d_compute(
attrs=conv2d_attrs,
)
+
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 6c9c927 Use shallow clone (#9864)
add 47d6eb2 [LLVM] LLVM codegen debug utilities (#9857)
No new revisions were ad
masahi merged pull request #9857:
URL: https://github.com/apache/tvm/pull/9857
--
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...
1 - 100 of 113 matches
Mail list logo
