chaozju opened a new issue #5804:
URL: https://github.com/apache/incubator-tvm/issues/5804
Is there a way to dump the Module to llvm ir into a `.s` file which can be
later built to standalone executable by `clang` ?
where I am:
- dump an output from python following toturial
masahi commented on pull request #5800:
URL: https://github.com/apache/incubator-tvm/pull/5800#issuecomment-643713382
Thanks @cbalint13
This is an automated message from the Apache Git Service.
To respond to the message,
masahi merged pull request #5800:
URL: https://github.com/apache/incubator-tvm/pull/5800
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.
masahi pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from a14a98e [TEST] Temporary disable fp16 type_as test for PyTorch
Frontend (#5799)
add 12558b6 [ONNX]
ceruleangu opened a new pull request #5803:
URL: https://github.com/apache/incubator-tvm/pull/5803
@antinucleon @vinx13 please review! ty:)
This is an automated message from the Apache Git Service.
To respond to the
siju-samuel commented on a change in pull request #5495:
URL: https://github.com/apache/incubator-tvm/pull/5495#discussion_r439784162
##
File path: tests/python/frontend/tflite/test_forward.py
##
@@ -1679,48 +1564,6 @@ def test_forward_squeeze():
tqchen commented on pull request #5801:
URL: https://github.com/apache/incubator-tvm/pull/5801#issuecomment-643707138
cc @vinx13 please to help manage this PR
This is an automated message from the Apache Git Service.
To
tqchen opened a new pull request #5802:
URL: https://github.com/apache/incubator-tvm/pull/5802
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
vinx13 commented on pull request #5758:
URL: https://github.com/apache/incubator-tvm/pull/5758#issuecomment-643704526
The tophub PR has been merged, please restart CI to try again
This is an automated message from the Apache
tqchen closed issue #5490:
URL: https://github.com/apache/incubator-tvm/issues/5490
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 #5792:
URL: https://github.com/apache/incubator-tvm/issues/5792#issuecomment-643703000
@zhiics would you be interested in taking a stab?
This is an automated message from the Apache Git Service.
To
tqchen closed issue #5798:
URL: https://github.com/apache/incubator-tvm/issues/5798
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 #5798:
URL: https://github.com/apache/incubator-tvm/issues/5798#issuecomment-643702952
This is due to the fact of us storinhg the tensor expression recursively.
There is an ongoing effort to move more of the parts into TIR, where the
sequential structure won't be
This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 8578096 [TIR][REFACTIR] Update TIR nodes std::string->String. (#5793)
add a14a98e [TEST] Temporary
tqchen merged pull request #5799:
URL: https://github.com/apache/incubator-tvm/pull/5799
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
cbalint13 opened a new pull request #5801:
URL: https://github.com/apache/incubator-tvm/pull/5801
This PR adds a long desired switch to enable quantization of ```nn.dense```.
By default ```nn.dense``` is not quantized to keep the current behaviour.
Thank You !
@tmoreau89
zhiics commented on a change in pull request #5770:
URL: https://github.com/apache/incubator-tvm/pull/5770#discussion_r439757500
##
File path: src/target/source/source_module.cc
##
@@ -152,8 +153,92 @@ runtime::Module DeviceSourceModuleCreate(
return runtime::Module(n);
}
anijain2305 commented on pull request #4277:
URL: https://github.com/apache/incubator-tvm/pull/4277#issuecomment-643655799
Closing as this is not relevant anymore given parallel efforts for improving
int8 conv schedules
anijain2305 closed pull request #4277:
URL: https://github.com/apache/incubator-tvm/pull/4277
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
cbalint13 opened a new pull request #5800:
URL: https://github.com/apache/incubator-tvm/pull/5800
This small PR optimize the GEMM (nn.dense) import via ONNX. It also makes
quantization decisions much better.
**Description**
A single ```Gemm``` operator from ONNX expands into
tqchen commented on pull request #5787:
URL: https://github.com/apache/incubator-tvm/pull/5787#issuecomment-643648813
Followup https://github.com/apache/incubator-tvm/pull/5799
This is an automated message from the Apache
tqchen opened a new pull request #5799:
URL: https://github.com/apache/incubator-tvm/pull/5799
cc @randxie @masahi
This is due to the problem of the code generator. The CI allows
non-TensorCore compatible GPUs(so we have more resources) in frontend tests to
enable maximum
tqchen commented on pull request #5779:
URL: https://github.com/apache/incubator-tvm/pull/5779#issuecomment-643646043
I don't think we disagree in many cases.
Let me try to rephrase @t-vi's the dtype part of the arguments:
- F0: NLP and quantization worklaods needs mixed
tqchen merged pull request #5793:
URL: https://github.com/apache/incubator-tvm/pull/5793
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.
tqchen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 59f5cbe support aten::type_as in the pytorch frontend (#5787)
add 8578096 [TIR][REFACTIR] Update TIR
junrushao1994 commented on pull request #5787:
URL: https://github.com/apache/incubator-tvm/pull/5787#issuecomment-643641262
Hey the master CI failure seems relevant to this PR. Could you take a look?
Thanks!
This is an
junrushao1994 commented on pull request #5796:
URL: https://github.com/apache/incubator-tvm/pull/5796#issuecomment-643641148
The CI failure is not relevant
This is an automated message from the Apache Git Service.
To respond
t-vi edited a comment on pull request #5779:
URL: https://github.com/apache/incubator-tvm/pull/5779#issuecomment-643601161
Sorry, but the dtypes discussion isn't for me. Working on NLP models like
BERT, if I have to argue that non-fp32 inputs are important, TVM is not a good
choice for
junrushao1994 commented on pull request #5740:
URL: https://github.com/apache/incubator-tvm/pull/5740#issuecomment-643640442
A3 is done over the last commits
This is an automated message from the Apache Git Service.
To
KnowingNothing opened a new issue #5798:
URL: https://github.com/apache/incubator-tvm/issues/5798
This is an extreme case in which tvm will give Segmentation Fault.
```py
import tvm
def test1(repeat=10):
A = tvm.te.placeholder([4, 4])
Output = tvm.te.compute([4,
FrozenGene commented on pull request #5753:
URL: https://github.com/apache/incubator-tvm/pull/5753#issuecomment-643614178
> @FrozenGene please let us know when it is ready for review
sure. i will update the status inside this pr and will notify you after
completion.
t-vi edited a comment on pull request #5779:
URL: https://github.com/apache/incubator-tvm/pull/5779#issuecomment-643601161
Sorry, but the dtypes discussion isn't for me. Working on NLP models like
BERT, if I have to argue that non-fp32 inputs are important, TVM is not a good
choice for
t-vi commented on pull request #5779:
URL: https://github.com/apache/incubator-tvm/pull/5779#issuecomment-643601161
Sorry, but the dtypes discussion isn't for me. Working on NLP models like
BERT, if I have to argue that non-fp32 inputs are important, TVM is not a good
choice for that
lixiaoquan opened a new pull request #5797:
URL: https://github.com/apache/incubator-tvm/pull/5797
A function may refer to other resources in the same module, so keep
the content of original module when infering a function.
cc @mbaret @mbrookhart @comaniac
junrushao1994 commented on a change in pull request #5601:
URL: https://github.com/apache/incubator-tvm/pull/5601#discussion_r439716507
##
File path: tests/python/unittest/test_tir_transform_bf16_legalize.py
##
@@ -0,0 +1,152 @@
+# Licensed to the Apache Software Foundation
35 matches
Mail list logo