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 8c125ca [RPC] Take PageAllocator out of MinRPCServer, make it
template parameter (#10219)
add 77fcd14 Skip
masahi merged pull request #10231:
URL: https://github.com/apache/tvm/pull/10231
--
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.
areusch pushed a change to branch ci-docker-staging
in repository https://gitbox.apache.org/repos/asf/tvm.git.
discard e6754d3 Validate Docker images using 20220204-110603-89aa4fe28
updating TF to 2.6.
discard 89aa4fe
driazati opened a new pull request #10231:
URL: https://github.com/apache/tvm/pull/10231
Since this test is launched in a thread, errors aren't propagated up to the
main thread and thusly pytest doesn't detect / report them, so this test will
always succeed. Given that this single test
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 6dece18 Adding support for Hexagon User DMA Engine (#10217)
add 8c125ca [RPC] Take PageAllocator out of
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 c20cbc5 Fixes for follow up on PR #9631 (#10205)
add 6dece18 Adding support for Hexagon User DMA Engine (#10217)
jroesch merged pull request #10219:
URL: https://github.com/apache/tvm/pull/10219
--
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:
jroesch merged pull request #10217:
URL: https://github.com/apache/tvm/pull/10217
--
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:
anwang2009 closed pull request #10171:
URL: https://github.com/apache/tvm/pull/10171
--
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:
anwang2009 commented on pull request #10171:
URL: https://github.com/apache/tvm/pull/10171#issuecomment-1036764415
Closing this. Sweeps show that the relay opt level 3 is the right approach
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
junrushao1994 commented on pull request #10230:
URL: https://github.com/apache/tvm/pull/10230#issuecomment-1036760628
CC: @vinx13 @spectrometerHBH @jinhongyii
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
Lunderberg commented on a change in pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#discussion_r805056854
##
File path: include/tvm/tir/buffer.h
##
@@ -55,8 +55,22 @@ class BufferNode : public Object {
Var data;
/*! \brief data type in the content of the
Lunderberg commented on a change in pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#discussion_r805050148
##
File path: include/tvm/tir/function.h
##
@@ -91,11 +91,23 @@ class PrimFuncNode : public BaseFuncNode {
*/
Map buffer_map;
+ /*! \brief The
Lunderberg commented on a change in pull request #9727:
URL: https://github.com/apache/tvm/pull/9727#discussion_r805042526
##
File path: include/tvm/te/schedule.h
##
@@ -771,6 +835,36 @@ class Singleton : public IterVarRelation {
TVM_DEFINE_OBJECT_REF_METHODS(Singleton,
Lunderberg opened a new pull request #10229:
URL: https://github.com/apache/tvm/pull/10229
- Modified `tvm.contrib.nvcc.get_cuda_version` to return a
`(major,minor,release)` tuple rather than a float.
- Implemented `tvm.testing.requries_nvcc_version` decorator to specify the
Raghav-Chakravarthy opened a new pull request #10228:
URL: https://github.com/apache/tvm/pull/10228
Issue: Deprecation warning in test_op_qnn_conv2_transpose.py
Solution: Used latest interface to replace the old use case
--
This is an automated message from the Apache Git
masahi commented on pull request #10085:
URL: https://github.com/apache/tvm/pull/10085#issuecomment-1036612700
Hi @chiwwang, can you take a look at the CI problem?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
kparzysz-quic opened a new pull request #10227:
URL: https://github.com/apache/tvm/pull/10227
This file is included every time TVM is build, regardless of whether any
support for Hexagon is enabled or not. This refactoring is meant to remove
underlying assumptions about what features are
tmoreau89 commented on pull request #10226:
URL: https://github.com/apache/tvm/pull/10226#issuecomment-1036528082
CC @mbrookhart @jwfromm @tkonolige @leandron
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
tmoreau89 opened a new pull request #10226:
URL: https://github.com/apache/tvm/pull/10226
[End to end benchmarking](https://github.com/apache/tvm/pull/8858) was added
to the VM and graph executor in order to faithfully represent execution time of
a graph module in order to account for
tmoreau89 closed pull request #10225:
URL: https://github.com/apache/tvm/pull/10225
--
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:
tmoreau89 opened a new pull request #10225:
URL: https://github.com/apache/tvm/pull/10225
[End to end benchmarking](https://github.com/apache/tvm/pull/8858) was added
to the VM and graph executor in order to faithfully represent execution time of
a graph module in order to account for
masahi commented on issue #10223:
URL: https://github.com/apache/tvm/issues/10223#issuecomment-1036489478
Thanks, yes `conv2d_transpose` with group was only recently fixed and
supported in https://github.com/apache/tvm/pull/9465. I think we haven't
updated our PyTorch frontend to benefit
manupa-arm edited a comment on pull request #10224:
URL: https://github.com/apache/tvm/pull/10224#issuecomment-1036440951
This is blocked on #10022 and changes belonging to this PR is only in the
commit :
manupa-arm edited a comment on pull request #10224:
URL: https://github.com/apache/tvm/pull/10224#issuecomment-1036440951
This is blocked on #10022 and changes belonging to this PR is only in the
commit :
manupa-arm commented on pull request #10224:
URL: https://github.com/apache/tvm/pull/10224#issuecomment-1036440951
This is blocked on #10022
--
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
manupa-arm opened a new pull request #10224:
URL: https://github.com/apache/tvm/pull/10224
This commit mainly enables the USMP with CMSIS-NN codegen.
In order to do that, CMSIS-NN functions needed to contain BufferMaps. This
commit adds the necessary BufferMaps as well.
All the tests
adstraw commented on a change in pull request #10217:
URL: https://github.com/apache/tvm/pull/10217#discussion_r804804080
##
File path: src/runtime/hexagon/hexagon/hexagon_user_dma_instructions.h
##
@@ -0,0 +1,75 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF)
This is an automated email from the ASF dual-hosted git repository.
kparzysz pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 2ac6cfe Fix more ONNX URLs (#10220)
add c20cbc5 Fixes for follow up on PR #9631 (#10205)
No new revisions were
manupa-arm commented on a change in pull request #46:
URL: https://github.com/apache/tvm-rfcs/pull/46#discussion_r804789970
##
File path: rfcs/0046-module-based-model-runtime-for-aot.md
##
@@ -0,0 +1,348 @@
+# Module-based Model Runtime Interface for AOT
+
+- Feature Name:
kparzysz-quic merged pull request #10205:
URL: https://github.com/apache/tvm/pull/10205
--
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:
kparzysz-quic edited a comment on pull request #10205:
URL: https://github.com/apache/tvm/pull/10205#issuecomment-1036365244
In
kparzysz-quic commented on pull request #10205:
URL: https://github.com/apache/tvm/pull/10205#issuecomment-1036365244
In
https://github.com/apache/tvm/pull/10205/files#diff-699c269e31b524f44d36ad4cc38d82f0e6f7edd8de3a005d583976d09ecb8b9bR198,
you don't need to use `new`. You can
JCBrouwer opened a new issue #10223:
URL: https://github.com/apache/tvm/issues/10223
I'm trying to convert a PyTorch model which makes use of
`torch.nn.functional.conv_transpose2d` and am running into issues with my
converter to the corresponding `tvm.relay.op.nn.conv2d_transpose`
pfk-beta opened a new pull request #10222:
URL: https://github.com/apache/tvm/pull/10222
+ remove useless directives in sc…
This PR solves issue 10221
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
pfk-beta commented on issue #10221:
URL: https://github.com/apache/tvm/issues/10221#issuecomment-1036246866
Hi @pfk-beta,
You need to replace in `ubuntu_install_python.sh` python3.6 to python3.7.
Btw. in this script you can remove python-pip and python-dev, because they
reference
pfk-beta opened a new issue #10221:
URL: https://github.com/apache/tvm/issues/10221
Hello,
I'm trying to following this tutorial:
https://tvm.apache.org/docs/how_to/deploy_models/deploy_model_on_android.html
But I cannot build docker image based on ubuntu 16.04, because it need
d-smirnov commented on a change in pull request #8509:
URL: https://github.com/apache/tvm/pull/8509#discussion_r804599588
##
File path: python/tvm/script/tir/scope_handler.py
##
@@ -157,6 +158,53 @@ def setup_buffer_var(
context.update_symbol(name, self.buffer_var,
d-smirnov commented on a change in pull request #8509:
URL: https://github.com/apache/tvm/pull/8509#discussion_r804593816
##
File path: include/tvm/ir/module.h
##
@@ -349,6 +386,9 @@ class IRModuleNode : public Object {
*/
std::unordered_set import_set_;
friend class
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 8aeb722 fix an index out of bound problem of cache write block
(#10203)
add 2ac6cfe Fix more ONNX URLs (#10220)
masahi merged pull request #10220:
URL: https://github.com/apache/tvm/pull/10220
--
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:
ashutosh-arm commented on issue #10213:
URL: https://github.com/apache/tvm/issues/10213#issuecomment-1036032779
@Mousius @grant-arm for visibility.
--
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
yzh119 edited a comment on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1035993696
> Looks like the perf improvement isn't very much? Only when n = 4 the
shuffle-down implementation is better than the shared memory implementation 樂
My typo, I have
yzh119 edited a comment on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1034535980
Sure, below is the measured time of the kernel:
```python
@T.prim_func
def reduce(a: T.handle, b: T.handle, n: T.int32) -> None:
A = T.match_buffer(a,
yzh119 edited a comment on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1035995100
> BTW do we have this requirement in the codebase now?
@MasterJH5574 yes there is a notion of `group_extent` and `reduce_extent`.
--
This is an automated message
yzh119 commented on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1035995100
> BTW do we have this requirement in the codebase now?
@MasterJH5574 yes there is a notion of `group_extent` and `reduce_extent`.
--
This is an automated message from the
yzh119 commented on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1035993696
@MasterJH5574 Ah I think n=4 is the only case shuffle-down worse than shared
memory.
Another benefit of using shuffle-down is reducing the shared memory usage
thus increasing
MasterJH5574 commented on pull request #10207:
URL: https://github.com/apache/tvm/pull/10207#issuecomment-1035982910
Interesting. Looks like the perf improvement isn't very much? Only when `n =
4` the shuffle-down implementation is better than the shared memory
implementation 樂
>
48 matches
Mail list logo