mehrdadh closed pull request #9026:
URL: https://github.com/apache/tvm/pull/9026
--
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:
huajsj commented on a change in pull request #9012:
URL: https://github.com/apache/tvm/pull/9012#discussion_r711431026
##
File path: src/runtime/logging.cc
##
@@ -166,10 +167,118 @@ namespace tvm {
namespace runtime {
namespace detail {
+namespace {
+constexpr const char*
spectrometerHBH opened a new pull request #9041:
URL: https://github.com/apache/tvm/pull/9041
This PR is part of the TensorIR upstreaming effort (#7527), which adds the
following schedule primitives:
decompose_reduction
Co-authored-by: Junru Shao junrushao1...@gmail.com
Raghav-Chakravarthy opened a new pull request #9040:
URL: https://github.com/apache/tvm/pull/9040
[Issue]
While reviewing the tvm code, I noticed some naming convention issues
in the diag_ctx_ and current_func variables.
Variable current_func should be current_func_ because it
tmoreau89 commented on issue #8838:
URL: https://github.com/apache/tvm/issues/8838#issuecomment-922117723
And thanks to @arangasa , https://github.com/apache/tvm/pull/9028 has been
merged! `com.microsoft.QLinearSigmoid` is now supported.
--
This is an automated message from the Apache
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 54484c2 Add bindings for StaticMemoryPlan and DensePackAttrs (#9034)
add a65bf34 [ONNX][#8838] QLinearSigmoid
masahi merged pull request #9028:
URL: https://github.com/apache/tvm/pull/9028
--
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:
mehrdadh commented on pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#issuecomment-922114683
@gromero thanks for the review. I addressed the comments. PTAL.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711376734
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711376548
##
File path: python/tvm/micro/project.py
##
@@ -101,7 +101,26 @@ def __init__(self, api_client):
if not self._info["is_template"]:
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711376357
##
File path: tests/micro/zephyr/test_zephyr.py
##
@@ -67,6 +71,7 @@ def _make_session(temp_dir, zephyr_board, west_cmd, mod,
build_config):
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711374226
##
File path: tests/micro/zephyr/conftest.py
##
@@ -53,6 +56,24 @@ def zephyr_boards() -> dict:
ZEPHYR_BOARDS = zephyr_boards()
+def qemu_boards(board:
gromero commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711374230
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
AndrewZhaoLuo opened a new pull request #9039:
URL: https://github.com/apache/tvm/pull/9039
Featuring: black magic from shape_funcs
--
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
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711369451
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
mehrdadh commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711367607
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
gromero commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711358099
##
File path: python/tvm/micro/project.py
##
@@ -101,7 +101,26 @@ def __init__(self, api_client):
if not self._info["is_template"]:
gromero commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711350143
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
masahi merged pull request #9034:
URL: https://github.com/apache/tvm/pull/9034
--
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 9100a3c [microTVM][autoTVM] Follow up fixes to #9003 (#9018)
add 54484c2 Add bindings for StaticMemoryPlan and
vinx13 edited a comment on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-922087058
--
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
vinx13 commented on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-922087058
Alternatively we can try creating a local array without deep copy
```
loc_args = []
for arg in args:
if arg is None:
loc_args.append(empty_array)
else:
echuraev commented on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-922084902
> Did you try set `args=None` or set all elements to None (the predefined
arrays are also converted to ndarray from numpy array here in each popen
workers locally, which should
tqchen commented on a change in pull request #9016:
URL: https://github.com/apache/tvm/pull/9016#discussion_r711350425
##
File path: src/arith/pattern_match.h
##
@@ -210,6 +210,73 @@ class PVar : public Pattern> {
mutable bool filled_{false};
};
+/*!
+ * \brief Wrapper
mbs-octoml commented on pull request #9038:
URL: https://github.com/apache/tvm/pull/9038#issuecomment-922072322
Oh, and this is on top of #9012 which is still in-flight, sorry about that.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
mbs-octoml commented on pull request #9038:
URL: https://github.com/apache/tvm/pull/9038#issuecomment-922071624
@electriclilies @mikepapadim @jroesch Here's part 1. Things to watch out for:
- How to I may device_copy.cc branch from context_analysis.cc?
- Did I go too far with device
masahi commented on issue #9037:
URL: https://github.com/apache/tvm/issues/9037#issuecomment-922068729
If you are asking about PyTorch SiLU, it is already supported.
https://github.com/apache/tvm/pull/8753
--
This is an automated message from the Apache Git Service.
To respond to the
masahi closed issue #9037:
URL: https://github.com/apache/tvm/issues/9037
--
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:
mbs-octoml opened a new pull request #9038:
URL: https://github.com/apache/tvm/pull/9038
Currently LowerTEPass (backend/te_compiler.cc) is a 'special' pass because it
depends on a side-input DeviceMap. We'd like to remove that side-input, and
instead recover the Device (and,
areusch commented on a change in pull request #9026:
URL: https://github.com/apache/tvm/pull/9026#discussion_r711327235
##
File path: apps/microtvm/zephyr/template_project/microtvm_api_server.py
##
@@ -59,44 +60,13 @@
# Data structure to hold the information
areusch commented on a change in pull request #9012:
URL: https://github.com/apache/tvm/pull/9012#discussion_r711268051
##
File path: include/tvm/runtime/logging.h
##
@@ -409,6 +416,68 @@ inline bool DebugLoggingEnabled() {
return state == 1;
}
+/*! \brief Helpers for \p
areusch commented on a change in pull request #9012:
URL: https://github.com/apache/tvm/pull/9012#discussion_r711264402
##
File path: include/tvm/runtime/logging.h
##
@@ -409,6 +416,68 @@ inline bool DebugLoggingEnabled() {
return state == 1;
}
+/*! \brief Helpers for \p
comaniac commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711256865
##
File path: Jenkinsfile
##
@@ -201,192 +205,240 @@ def unpack_lib(name, libs) {
}
stage('Build') {
- parallel 'BUILD: GPU': {
-node('GPUBUILD') {
mikepapadim commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711248909
##
File path: tests/scripts/git_check_tree.sh
##
@@ -0,0 +1,40 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+#
mikepapadim commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711248479
##
File path: Jenkinsfile
##
@@ -201,192 +205,240 @@ def unpack_lib(name, libs) {
}
stage('Build') {
- parallel 'BUILD: GPU': {
-
mikepapadim commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711248257
##
File path: Jenkinsfile
##
@@ -201,192 +205,240 @@ def unpack_lib(name, libs) {
}
stage('Build') {
- parallel 'BUILD: GPU': {
-
comaniac commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711241743
##
File path: Jenkinsfile
##
@@ -148,11 +149,15 @@ stage('Prepare') {
}
}
-stage("Sanity Check") {
+stage('Sanity Check') {
timeout(time: max_time,
areusch commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711223304
##
File path: tests/scripts/git_check_tree.sh
##
@@ -0,0 +1,40 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or
This is an automated email from the ASF dual-hosted git repository.
areusch pushed a change to branch ci-docker-staging
in repository https://gitbox.apache.org/repos/asf/tvm.git.
discard 7a62a90 Test docker images tlcpackstaging/*:20210823-232401-607bb9492
add 596ff5c [AutoTVM] Use
tqchen edited a comment on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921820526
Please allow me to explain the overall rationale here, in particular over
the term "constraint"
- C0: On one hand, we want a "default" memory to be generically accessible
tqchen commented on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921945722
Thanks @manupa-arm. I agree that putting TVMBAW as a peer to heap is not
right(that was meant as an example to demonstrate the viewpoint. I do not
necessary want to enforce heap as a
comaniac commented on a change in pull request #9031:
URL: https://github.com/apache/tvm/pull/9031#discussion_r711204273
##
File path: Jenkinsfile
##
@@ -42,41 +42,42 @@
// Hashtag in the source to build current CI docker builds
//
//
+import
manupa-arm edited a comment on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921934672
Thanks @tqchen for the explanation of two viewpoints of how we could see the
constraints.
I do agree that we should put more constraints progressively to guide the
manupa-arm commented on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921934672
Thanks @tqchen for the explanation of two viewpoints of how we could see the
constraints.
I do agree that we should put more constraints progressively to guide the
compiler in
This is an automated email from the ASF dual-hosted git repository.
areusch pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 1fd8f61 [Hexagon] Don't use {} initialization with FastRPC structures
(#9033)
add 9100a3c [microTVM][autoTVM]
areusch merged pull request #9018:
URL: https://github.com/apache/tvm/pull/9018
--
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:
tqchen edited a comment on pull request #23:
URL: https://github.com/apache/tvm-rfcs/pull/23#issuecomment-921900930
Thanks @manupa-arm , sorry for the delay. I think the proposal(of adding
annotations) looks good.
As a formal procedure for IR change. It would be great to have an
ashutosh-arm commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r711172103
##
File path: src/relay/backend/contrib/cmsisnn/tir_to_runtime.cc
##
@@ -0,0 +1,138 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
ashutosh-arm commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r711171886
##
File path: src/relay/backend/contrib/cmsisnn/relay_to_tir.cc
##
@@ -0,0 +1,140 @@
+
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
tqchen commented on pull request #23:
URL: https://github.com/apache/tvm-rfcs/pull/23#issuecomment-921900930
Thanks @manupa-arm . I think the proposal(of adding annotations) looks good.
Perhaps we can change the title to
adding annotation fields to the Allocation, and discuss the
tmoreau89 commented on issue #8838:
URL: https://github.com/apache/tvm/issues/8838#issuecomment-921889564
@quic-sanirudh please go ahead and pick
`com.microsoft.DynamicQuantizeMatMul` given that no one else has claimed, it. I
can mark you down for it.
--
This is an automated message
jinfagang opened a new issue #9037:
URL: https://github.com/apache/tvm/issues/9037
Will it considered to support?
--
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
vinx13 edited a comment on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-921848437
I see. If I understand correctly, the issue is when popen worker got a copy
of `args` here
https://github.com/apache/tvm/blob/main/python/tvm/exec/popen_worker.py#L77 and
Alex-grovety commented on issue #8717:
URL: https://github.com/apache/tvm/issues/8717#issuecomment-921856470
If I understand correctly it will be
```
elif layout == "NHWC":
assert kernel_layout == "HWOI"
is_aarch64 = topi.arm_cpu.arm_utils.is_aarch64_arm()
if
vinx13 edited a comment on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-921848437
I see. If I understand correctly, the issue is when popen worker got a copy
of `args` here
https://github.com/apache/tvm/blob/main/python/tvm/exec/popen_worker.py#L77 and
vinx13 commented on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-921848437
I see. If I understand correctly, the issue is when popen worker got a copy
of `args` here
https://github.com/apache/tvm/blob/main/python/tvm/exec/popen_worker.py#L77 can
call
tqchen edited a comment on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921820526
Please allow me to explain the overall rationale here, in particular over
the term "constraint"
- C0: On one hand, we want a "default" memory to be generically accessible
tqchen commented on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921820526
Please allow me to explain the overall rationale here, in particular over
the term "constraint"
- C0: On one hand, we want a "default" memory to be generically accessible
(per
mbs-octoml commented on pull request #9012:
URL: https://github.com/apache/tvm/pull/9012#issuecomment-921803740
@huajsj @areusch PTAL, and CI is green :-)
--
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
quic-sanirudh commented on issue #8838:
URL: https://github.com/apache/tvm/issues/8838#issuecomment-921793403
@tmoreau89 Thanks for your help in reviewing the AveragePool op support
quickly. Shall I pick up the com.microsoft.DynamicQuantizeMatMul op next to
work on, if no one else has
ashutosh-arm commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r711048086
##
File path: src/relay/backend/contrib/cmsisnn/relay_to_tir.cc
##
@@ -0,0 +1,140 @@
+
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
ashutosh-arm commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r711047932
##
File path: src/relay/backend/contrib/cmsisnn/relay_to_tir.cc
##
@@ -0,0 +1,140 @@
+
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
echuraev commented on pull request #9032:
URL: https://github.com/apache/tvm/pull/9032#issuecomment-921787831
> Empty arrays created in timed_rpc_run should be destroy after running the
program in the device.
Who will guarantee that the arrays will be destroyed? We have a global
manupa-arm edited a comment on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921782373
@tqchen ,
> To say it in another way, we cannot say that "global" definitely mean no
stack allocation.
The current issue is in the device "CPU" && 'global'
manupa-arm commented on issue #9022:
URL: https://github.com/apache/tvm/issues/9022#issuecomment-921782373
@tqchen ,
> To say it in another way, we cannot say that "global" definitely mean no
stack allocation.
The current issue is in the device "CPU" && 'global' certainly
sergey-grovety commented on a change in pull request #8990:
URL: https://github.com/apache/tvm/pull/8990#discussion_r711021647
##
File path: apps/microtvm/reference-vm/zephyr/base-box/base_box_test.sh
##
@@ -37,3 +37,5 @@ if [ $board == "stm32f746xx" ]; then
else
pytest
arangasa edited a comment on pull request #9028:
URL: https://github.com/apache/tvm/pull/9028#issuecomment-921535996
> The de-quantize change did cause those tests to fail in CI. I think we
need a better solution there, but otherwise, it looks good.
Thank you, @mbrookhart ,
PhilippvK opened a new issue #9036:
URL: https://github.com/apache/tvm/issues/9036
I recently ran into segmentation faults working with rather large BYOC
subgraphs and the AoT executor for MicroTVM. It seems to be related to a BYOC
subgraph with more than one output, as I was able to
jiangjiajun closed pull request #9035:
URL: https://github.com/apache/tvm/pull/9035
--
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:
jiangjiajun opened a new pull request #9035:
URL: https://github.com/apache/tvm/pull/9035
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
Mousius commented on pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#issuecomment-921682639
Hi @areusch, I acknowledge your questions but do we need to block the PR on
them rather than moving forwards with the integration? I'm sure @ashutosh-arm
will honourably follow up
Mousius commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r710931393
##
File path: src/relay/backend/contrib/cmsisnn/tir_to_runtime.cc
##
@@ -0,0 +1,138 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ *
Mousius commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r710930215
##
File path: src/relay/backend/contrib/cmsisnn/relay_to_tir.cc
##
@@ -0,0 +1,140 @@
+
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ *
Mousius commented on a change in pull request #8951:
URL: https://github.com/apache/tvm/pull/8951#discussion_r710929324
##
File path: src/relay/backend/contrib/cmsisnn/codegen_cmsisnn.cc
##
@@ -16,190 +16,36 @@
* specific language governing permissions and limitations
*
Mousius commented on a change in pull request #28:
URL: https://github.com/apache/tvm-rfcs/pull/28#discussion_r710900304
##
File path: rfcs/0028-command-line-registry-composition.md
##
@@ -0,0 +1,108 @@
+- Feature Name: Command Line Composition from Internal Registry
+- Start
Mousius commented on a change in pull request #28:
URL: https://github.com/apache/tvm-rfcs/pull/28#discussion_r710898391
##
File path: rfcs/0028-command-line-registry-composition.md
##
@@ -0,0 +1,108 @@
+- Feature Name: Command Line Composition from Internal Registry
+- Start
rafzi opened a new pull request #9034:
URL: https://github.com/apache/tvm/pull/9034
- DensePackAttrs was missing
- StaticMemoryPlan (and containing StorageInfo) were not constructible from
python. However, this is useful when doing custom memory planning.
--
This is an automated
arangasa commented on pull request #9028:
URL: https://github.com/apache/tvm/pull/9028#issuecomment-921535996
> The de-quantize change did cause those tests to fail in CI. I think we
need a better solution there, but otherwise, it looks good.
Thank you, @mbrookhart , @anwang2009 ,
78 matches
Mail list logo
