Maximilianxu opened a new issue #7758:
URL: https://github.com/apache/tvm/issues/7758
Hi
I want to deploy the BERT-base model on an Android phone. One of its params
has shape (30522, 768) with dtype float32, the RPC connection will be reset
each time I allocate this array.
FrozenGene commented on a change in pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#discussion_r602653607
##
File path: src/arith/iter_affine_map.cc
##
@@ -459,27 +665,107 @@ class IterMapRewriter : public ExprMutator {
}
};
+/*! \brief An internal
jwfromm commented on a change in pull request #7747:
URL: https://github.com/apache/tvm/pull/7747#discussion_r602647038
##
File path: tests/python/frontend/onnx/test_forward.py
##
@@ -4278,6 +4278,33 @@ def test_wrong_input():
relay.frontend.from_onnx(model,
spectrometerHBH edited a comment on pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#issuecomment-808608709
> Thanks @spectrometerHBH There seems to be three places where the
ExprComplexity is involved.
>
> *
spectrometerHBH commented on pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#issuecomment-808608709
> Thanks @spectrometerHBH There seems to be three places where the
ExprComplexity is involved.
>
> *
masahi commented on pull request #7438:
URL: https://github.com/apache/tvm/pull/7438#issuecomment-808604311
@electriclilies Please try kick again, the flaky issue should be fixed now
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
This is an automated email from the ASF dual-hosted git repository.
junrushao pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 14f829a [docs] Getting Started with TVM: AutoTVM and Matrix Multiply
(#7643)
add 474bc4e [RUNTIME][WEB]
junrushao1994 merged pull request #7750:
URL: https://github.com/apache/tvm/pull/7750
--
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.
For queries about this service,
junrushao1994 commented on pull request #7750:
URL: https://github.com/apache/tvm/pull/7750#issuecomment-808603577
Thanks @tqchen @tkonolige! I am going to merge it in :-)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
MasterJH5574 commented on a change in pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#discussion_r602639137
##
File path: src/arith/iter_affine_map.cc
##
@@ -161,59 +160,116 @@ class IterMarkSplitCollector {
}
};
-// Rewriter to rewrite PrimExpr to
tqchen edited a comment on pull request #7757:
URL: https://github.com/apache/tvm/pull/7757#issuecomment-808602675
cc @rkimball @junrushao1994 @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
tqchen commented on pull request #7757:
URL: https://github.com/apache/tvm/pull/7757#issuecomment-808602675
cc @rkimball @junrushao1994 @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
tqchen opened a new pull request #7757:
URL: https://github.com/apache/tvm/pull/7757
Previously the proxy relies multiprocessing to start a new process. This PR
switches that to use PopenWorker. We will send a worker function that:
- Starts a new thread with the latest state in the
junrushao1994 opened a new pull request #7756:
URL: https://github.com/apache/tvm/pull/7756
This PR brings back the stack size optimization:
- https://github.com/dmlc/dmlc-core/pull/613/
- https://github.com/dmlc/dmlc-core/pull/615/
It also brings back the error message showing
jroesch commented on pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#issuecomment-808590330
The last commit resolves all the outstanding comments, we will wait to be on
3.7 to use type annotations and I added doc comments and clarifications.
--
This is an automated
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602630766
##
File path: src/runtime/library_module.cc
##
@@ -106,16 +106,16 @@ Module LoadModuleFromBinary(const std::string& type_key,
dmlc::Stream* stream) {
if
areusch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602627326
##
File path: python/tvm/runtime/vm.py
##
@@ -299,12 +300,16 @@ class VirtualMachine(object):
POOLED_ALLOCATOR = 2
def __init__(self, exe,
mbrookhart opened a new pull request #7755:
URL: https://github.com/apache/tvm/pull/7755
cc @jwfromm
--
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.
For queries about
This is an automated email from the ASF dual-hosted git repository.
jroesch pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from f74dac7 Bump ci-cpu and ci-arm container versions. (#7745)
add 14f829a [docs] Getting Started with TVM: AutoTVM
jroesch merged pull request #7643:
URL: https://github.com/apache/tvm/pull/7643
--
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.
For queries about this service, please
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602618588
##
File path: src/runtime/vm/executable.cc
##
@@ -476,8 +484,19 @@ void LoadHeader(dmlc::Stream* strm) {
}
runtime::Module Executable::Load(const
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602618450
##
File path: src/runtime/vm/executable.cc
##
@@ -476,8 +484,19 @@ void LoadHeader(dmlc::Stream* strm) {
}
runtime::Module Executable::Load(const
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602616950
##
File path: src/runtime/vm/executable.cc
##
@@ -476,8 +484,19 @@ void LoadHeader(dmlc::Stream* strm) {
}
runtime::Module Executable::Load(const
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602616808
##
File path: src/runtime/library_module.cc
##
@@ -99,6 +99,28 @@ void InitContextFunctions(std::function
fgetsymbol) {
#undef TVM_INIT_CONTEXT_FUNC
}
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602616766
##
File path: src/runtime/vm/executable.cc
##
@@ -74,6 +76,12 @@ PackedFunc Executable::GetFunction(const std::string& name,
const
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602616507
##
File path: python/tvm/runtime/vm.py
##
@@ -299,12 +300,16 @@ class VirtualMachine(object):
POOLED_ALLOCATOR = 2
def __init__(self, exe,
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602615151
##
File path: python/tvm/runtime/module.py
##
@@ -269,24 +269,31 @@ def _collect_dso_modules(self):
return
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602613654
##
File path: include/tvm/runtime/vm/executable.h
##
@@ -125,9 +133,17 @@ class Executable : public ModuleNode {
* \brief Get the `lib` module in an
jroesch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602611761
##
File path: src/runtime/library_module.cc
##
@@ -99,6 +99,29 @@ void InitContextFunctions(std::function
fgetsymbol) {
#undef TVM_INIT_CONTEXT_FUNC
}
areusch commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602604080
##
File path: include/tvm/runtime/vm/executable.h
##
@@ -63,6 +63,14 @@ class Executable : public ModuleNode {
*/
PackedFunc GetFunction(const
tkonolige commented on a change in pull request #7746:
URL: https://github.com/apache/tvm/pull/7746#discussion_r602411072
##
File path: src/runtime/library_module.cc
##
@@ -99,6 +99,29 @@ void InitContextFunctions(std::function
fgetsymbol) {
#undef TVM_INIT_CONTEXT_FUNC
}
mbrookhart opened a new pull request #7754:
URL: https://github.com/apache/tvm/pull/7754
Also fixes partitioning functions whose bodies are matched fuzzily.
cc @jroesch @masahi @slyubomirsky @ekalda
Also see discuss forums here:
jroesch commented on pull request #7503:
URL: https://github.com/apache/tvm/pull/7503#issuecomment-808521165
@schell sorry I got really busy with some parallel work at OctoML and I
didn't realize anyone was blocked on this, will try and land over the weekend!
sorry for the latency here
areusch commented on pull request #7557:
URL: https://github.com/apache/tvm/pull/7557#issuecomment-808458855
Filed https://github.com/zephyrproject-rtos/zephyr/issues/33745 to track the
underlying Zephyr issue
--
This is an automated message from the Apache Git Service.
To respond to
trevor-m opened a new pull request #7753:
URL: https://github.com/apache/tvm/pull/7753
Squeeze takes a list of ints, but keras frontend had a case where it used a
single int which caused the following error:
```
TVMError: In function relay.op._make.squeeze: error while converting
areusch commented on pull request #7557:
URL: https://github.com/apache/tvm/pull/7557#issuecomment-808411220
@mdw-octoml I commented in a resolved thread expecting it to re-open, but
unsure if it did. please take a look at that one referenced in my latest review.
--
This is an automated
areusch commented on a change in pull request #7557:
URL: https://github.com/apache/tvm/pull/7557#discussion_r602488241
##
File path: python/tvm/micro/contrib/zephyr.py
##
@@ -650,10 +650,10 @@ def popen_kwargs(self):
env = dict(os.environ)
env["ZEPHYR_BASE"]
This is an automated email from the ASF dual-hosted git repository.
jwfromm pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from f449070 [Bugfix] Fix usages of logging-related macros (#7748)
add f74dac7 Bump ci-cpu and ci-arm container
areusch commented on pull request #7557:
URL: https://github.com/apache/tvm/pull/7557#issuecomment-808377621
I thnk this is your error now:
```
conv2d: requires that `0`, the input channels (0) divided by groups (1),
must match the input channels of the weight `1`, where the
jwfromm merged pull request #7745:
URL: https://github.com/apache/tvm/pull/7745
--
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.
For queries about this service, please
tkonolige commented on a change in pull request #7518:
URL: https://github.com/apache/tvm/pull/7518#discussion_r602419800
##
File path: src/relay/backend/graph_runtime_codegen.cc
##
@@ -181,23 +186,56 @@ class GraphOpNode : public GraphNode {
const std::string
areusch commented on pull request #7745:
URL: https://github.com/apache/tvm/pull/7745#issuecomment-808367289
staging run:
https://ci.tlcpack.ai/blue/organizations/jenkins/tvm/detail/ci-docker-staging/68/pipeline
for posterity
--
This is an automated message from the Apache Git Service.
areusch commented on pull request #7628:
URL: https://github.com/apache/tvm/pull/7628#issuecomment-808366585
sure. I don't believe it is included in #7745
--
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 issue #7751:
URL: https://github.com/apache/tvm/issues/7751#issuecomment-808358438
@areusch they are pinned because we want to make sure no cache is
needed(e.g. if we change version there will be automatic update process, but
won't reinstall other wise).
Note
areusch commented on pull request #7649:
URL: https://github.com/apache/tvm/pull/7649#issuecomment-808340791
just to clarify--seems like we have two options:
1. merge #7745 today, then rebuild ci-cpu and ci-arm again and merge a new
PR to bump them again
2. rebuild ci-cpu and ci-arm
tkonolige commented on a change in pull request #7750:
URL: https://github.com/apache/tvm/pull/7750#discussion_r602402636
##
File path: include/tvm/runtime/logging.h
##
@@ -55,6 +66,15 @@
#define TVM_USE_LIBBACKTRACE 0
#endif
+/*!
+ * \brief Whether or not customize the
leandron commented on issue #7751:
URL: https://github.com/apache/tvm/issues/7751#issuecomment-808327665
> yeah I also hit this problem. I think we need to fix the CI image update
process so it is not such a burden. My opinion is we should implement one
method to include dependencies in
areusch edited a comment on issue #7751:
URL: https://github.com/apache/tvm/issues/7751#issuecomment-808314918
yeah I also hit this problem. I think we need to fix the CI image update
process so it is not such a burden. My opinion is we should implement one
method to include dependencies
areusch commented on issue #7751:
URL: https://github.com/apache/tvm/issues/7751#issuecomment-808314918
yeah I also hit this problem. I think we need to fix the CI image update
process so it is not such a burden. My opinion is we should implement one
method to include dependencies in TVM
tqchen commented on a change in pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#discussion_r602342033
##
File path: src/arith/iter_affine_map.cc
##
@@ -459,27 +665,107 @@ class IterMapRewriter : public ExprMutator {
}
};
+/*! \brief An internal struct to
tqchen commented on pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#issuecomment-808273920
--
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.
For queries
tqchen commented on a change in pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#discussion_r602342033
##
File path: src/arith/iter_affine_map.cc
##
@@ -459,27 +665,107 @@ class IterMapRewriter : public ExprMutator {
}
};
+/*! \brief An internal struct to
tqchen commented on pull request #7752:
URL: https://github.com/apache/tvm/pull/7752#issuecomment-808248236
Thanks @spectrometerHBH . @MasterJH5574 @hzfan @Hzfengsy please also help to
take a look when you have time
--
This is an automated message from the Apache Git Service.
To respond
spectrometerHBH opened a new pull request #7752:
URL: https://github.com/apache/tvm/pull/7752
Update on previous PR https://github.com/apache/tvm/pull/6667
Enhance the detect_iter_map utility to be able to detect split patterns with
predicates.
cc @tqchen
--
This is an
tqchen edited a comment on pull request #7750:
URL: https://github.com/apache/tvm/pull/7750#issuecomment-808231650
cc @junrushao1994 @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
tqchen commented on issue #7751:
URL: https://github.com/apache/tvm/issues/7751#issuecomment-808237112
Thanks @leandron Because these two deps can be updated more frequently(and
ci image update cycle is slower), we choose to use ci_setup script to install
them instead of hard baking the
tqchen closed issue #7734:
URL: https://github.com/apache/tvm/issues/7734
--
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.
For queries about this service, please contact
tqchen commented on issue #7734:
URL: https://github.com/apache/tvm/issues/7734#issuecomment-808235720
Thanks @majercakdavid Please make sure that wasm_runtime tvmjs_support are
linked in your wasm. It is interesting that the partcular function is not
registered, as it is done in
leandron opened a new issue #7751:
URL: https://github.com/apache/tvm/issues/7751
There are some dependencies that are inconsistently declared within Docker
setup scripts and some CI test scripts. The cases I noticed are declared within
`docker/install/ubuntu_install_python_package.sh`,
tqchen commented on pull request #7750:
URL: https://github.com/apache/tvm/pull/7750#issuecomment-808231650
cc @tkonolige @junrushao1994
--
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 opened a new pull request #7750:
URL: https://github.com/apache/tvm/pull/7750
The log(info) stops working for web runtime due to the usage of time
function.
- Introduce TVM_LOG_CUSTOMIZE anf TVM_LOG_STACK_TRACE.
- Reorganize the log customization code to wasm_runtime(so
tqchen commented on issue #7749:
URL: https://github.com/apache/tvm/issues/7749#issuecomment-808174384
THanks @Elonaever, please ope a new thread on https://discuss.tvm.apache.org/
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
tqchen closed issue #7749:
URL: https://github.com/apache/tvm/issues/7749
--
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.
For queries about this service, please contact
tqchen commented on pull request #7653:
URL: https://github.com/apache/tvm/pull/7653#issuecomment-808173098
oops. sorry there is a merge conflict due to another renaming PR. @areusch
can you rebase. will prioritize merging this
--
This is an automated message from the Apache Git
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 fbfeee4 [Refactor] Rename TVMContext to Device (#7721)
add f449070 [Bugfix] Fix usages of logging-related macros
tqchen commented on pull request #7748:
URL: https://github.com/apache/tvm/pull/7748#issuecomment-808159571
Thank you @junrushao1994 !
--
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 merged pull request #7748:
URL: https://github.com/apache/tvm/pull/7748
--
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.
For queries about this service, please
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 8c8971a [CI] Temp disable rust docs build (#7743)
add fbfeee4 [Refactor] Rename TVMContext to Device (#7721)
No
tqchen merged pull request #7721:
URL: https://github.com/apache/tvm/pull/7721
--
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.
For queries about this service, please
tqchen commented on pull request #7721:
URL: https://github.com/apache/tvm/pull/7721#issuecomment-808159289
Thanks @icemelon9 this PR is now merged!
--
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
u99127 commented on pull request #7649:
URL: https://github.com/apache/tvm/pull/7649#issuecomment-808153935
Hi @d-smirnov - looks like we have this request coming through.
https://github.com/apache/tvm/pull/7745 which should handle it when things
land.
--
This is an automated message
u99127 commented on pull request #7628:
URL: https://github.com/apache/tvm/pull/7628#issuecomment-808152945
@areusch - could this also get a new ci-cpu image ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
d-smirnov commented on pull request #7649:
URL: https://github.com/apache/tvm/pull/7649#issuecomment-808149895
Ping. @zhiics , @tqchen is there anything we can help to get this patch
landed?
--
This is an automated message from the Apache Git Service.
To respond to the message, please
liangfu commented on a change in pull request #7680:
URL: https://github.com/apache/tvm/pull/7680#discussion_r602141479
##
File path: apps/android_rpc/app/src/main/jni/CMakeLists.txt
##
@@ -0,0 +1,51 @@
+cmake_minimum_required(VERSION 3.4.1)
+
+add_library(
+
leeexyz commented on issue #7596:
URL: https://github.com/apache/tvm/issues/7596#issuecomment-807985380
> @leeexyz , could you please check my screenshot posted above? This is
definitely an issue that a variable named "tvm.contrib.cblas.matmul_packed"
generated in the c source file, which
75 matches
Mail list logo