Ecosyste.ms: OpenCollective
An open API service for software projects hosted on Open Collective.
github.com/llvm/torch-mlir
The Torch-MLIR project aims to provide first class support from the PyTorch ecosystem to the MLIR ecosystem.
https://github.com/llvm/torch-mlir
* Initial changes
* Fix up native functions
* Further fix decoupling
* Remove unnecessary ops...
1bde00c73d58d98e83d3a9d3f51d3495611067e8 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>* Save InputOutputAliases to TorchMlirComputation
* Implement GetResultShape for TorchMlirLow...
cca9fe126e4e8ada6423202770ea904bfd22c5a1 authored about 2 years ago by Henry Tu <[email protected]>615ff1d31c265e45d98562bcdf690fbf2d0be8e5 authored about 2 years ago by Antonio Kim <[email protected]>
* Changes as a result of the LTC TS backend decoupling
* Fix bugs in BackendImpl and codegen
...
c3b20e444c01d1884978f402da056cddf632171d authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>
* Codegen and build LTC lowering
* Add LazyShapeInference header
58338f79a1030a9eaa8a17c6a8cdd63a50be564e authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>* Create ltc_backend.md
* Added introduction and examples
* Added descriptions for files f...
61db88c9ecaa30ebcb99b818efbb54b83318b6fa authored about 2 years ago by Henry Tu <[email protected]>* Added JIT to MLIR lowering
Lowering to JIT is performed in a way similar to how it's done in ...
3e9b1cbd361fcc96ab2e0c6a03de48ff12f7a57c authored about 2 years ago by Henry Tu <[email protected]>* Add initial LTC backend skeleton
* Disable CI build and move TorchMLIRPyTorch.cmake
2f22e2ef40f2bcf61480d338474e2d1d9246daca authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>* Update native function definitions
* Add ops to support bert lowering
- Add empty_strided an...
d9aee0d7a709933915b753964896d968afa2beb2 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>a605fe279cc35bbeec478e2436fc3f63ac102d5e authored about 2 years ago by Henry Tu <[email protected]>
0cee0dc978f5c73a325a1110979456c4c6cb8fff authored about 2 years ago by Henry Tu <[email protected]>
* Xfail unsupported ops
* Register FuncDialect
* Include dynamic_ir in build
* Code reformat
...
* Replace CHECK_EQ with TORCH_CHECK_EQ
* Check value of TORCH_MLIR_USE_INSTALLED_PYTORCH during...
2c3b3606d058ea0e109907fe47db5992ea7eefe0 authored about 2 years ago by Henry Tu <[email protected]>* Remove unnecessary sed in autogen
* Remove .pyc files frrom VCS
425362263bb1ff6c163eb9d66681330168aa2b67 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>368963243ef992b430ddb432a0d87b666a364722 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>
0d16a916567884b07639831a8166ad0805e1aaf9 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>
* Add default device ordinal API
* Fix reference backend
e37891b9978d38952abb03f6c084e9adefa280b1 authored about 2 years ago by Jae Hoon (Antonio) Kim <[email protected]>- Update llvm-project pin to match main
de6c135dc30724bdaf9b63a9fced2c07f83930cb authored about 2 years ago by Antonio Kim <[email protected]>Signed-Off By: Phaneesh Barwaria <[email protected]>
8b5631d4c57606a8c8061e0c351ac211a2dee4e8 authored about 2 years ago by PhaneeshB <[email protected]>* Update buildAndTest.yml
test with fast-fail matrix builds
* Remove redundant and stateme...
db4a6991a08e23ad6bbc7eff4461dc2f2514e50a authored about 2 years ago by powderluv <[email protected]>Fixes https://github.com/llvm/torch-mlir/issues/1121
Signed-Off By: Vivek Khandelwal<vivek@nod-...
9a1203c8446affcbe084e3fee364bdcd89ad2862 authored about 2 years ago by Vivek Khandelwal <[email protected]>b389053653a10783b9b075f52436ce6cb055de72 authored about 2 years ago by Ahmed S. Taei <[email protected]>
This commit adds the decomposition for `aten.var.correction` op.
Signed-Off By: Vivek Khandelwa...
d386b8f9e504fbfd28dc5404a238b917b557e8a9 authored about 2 years ago by Vivek Khandelwal <[email protected]>This commit adds lowering of `aten.ge.int` op.
Signed-Off By: Vivek Khandelwal <vivek@nod-labs....
7247c6a3a76482f6c9c4589c8c4f42872e7a1e1a authored about 2 years ago by Vivek Khandelwal <[email protected]>
This commit fixes the shape calculation for:
1.) aten.mean.dim
2.) aten.var.dim
3.) aten.sum.dim...
- Includes a canonicalizer for `aten.add.t`needed for successfully lowering the shape function
...
b36a17c9d23b8378d7e22c87b6ba632142874eed authored about 2 years ago by Sean Silva <[email protected]>
3c9addf19cd5eae62562600294150534dc353eae authored about 2 years ago by Quinn Dawkins <[email protected]>
* [MHLO] Init MHLO basic Op Conversion
Co-authored-by: Bairen Yi <[email protected]>...
Add source builds and remove deprecated libtorch information.
e23fbc89f27aa8ad190ce0903eb28a7fa0795c19 authored about 2 years ago by powderluv <[email protected]>Follows existing conventions for unary operators.
e8f327cc0063dcc283eab723c1d34d7abf4ea940 authored about 2 years ago by Kevin Kiningham <[email protected]>See RFC https://github.com/llvm/torch-mlir/issues/999
Co-authored-by: Bairen Yi yibairen.byro...
44ead687721abdc41817d17cb9dd8dc255a860a4 authored about 2 years ago by Tanyo Kwok <[email protected]>* [MHLO] Add [un]squeeze op patterns
* Conform to llvm coding standard
* minor update
f50d7013cdae42863a4eff05e664d8b647d6f311 authored about 2 years ago by Tanyo Kwok <[email protected]>f424930a28781c80a8c9fca6a2447b340ca573dd authored about 2 years ago by powderluv <[email protected]>
* [MHLO] Init MHLO view like op patterns
See RFC: https://github.com/llvm/torch-mlir/issues/9...
b80ce79b9fd26247e935899ce65f5cc11ec3e346 authored about 2 years ago by Tanyo Kwok <[email protected]>See RFC: https://github.com/llvm/torch-mlir/issues/999
Co-authored-by: Bairen Yi yibairen.byr...
a02dbb2d5ead1fc28a910a0a7f2aad6e90779f4a authored about 2 years ago by Tanyo Kwok <[email protected]>
This commit adds verifiers to the ops `ToBuiltinTensorOp` and
`FromBuiltinTensorOp` that make s...
This enables building Pytorch from source in the CI.
The build should mostly hit the ccache.
R...
The unknown dtype case can come from RefineTypes.
c0ef192865c0cfcf565fb5fbe400597fbc707463 authored about 2 years ago by Sean Silva <[email protected]>
This reverts commit ad283c104339267fbe11cc7efefc02a6ad4d2d1e, since it's
causing nightly build ...
In the interest of merging upstream LLVM quickly, a previous patch
(7f08169) updated the torch-...
Co-authored-by: Bairen Yi <[email protected]>
Co-authored-by: Jiawei Wu <xremold@gma...
- Supports cases where the view op expands and collapses dims
simulataneously. This does not ha...
An upstream MLIR bug (that was recently fixed) caused the result to be
ignored for Region- and ...
21f905afbeee76acb086d502ab6ed6eba7b9423b authored over 2 years ago by Kevin Kiningham <[email protected]>
This patch still only supports a single indexing tensor.
c73a39e40a2b2c0a6dc88d1937c9f41642c95086 authored over 2 years ago by Quinn Dawkins <[email protected]>
This patch makes some rudimentary changes to torch-mlir's use of MLIR
Python bindings to work w...
This commit adds the support for negative dim cases for `aten.cat`,
`aten.slice.Tensor` and `ate...
This commit adds the canonicalization pattern for the `prim.ListUnpack` op.
Signed-Off By: Vive...
4c25878e643a0806a93e34f6f0f16452cfd7e75a authored over 2 years ago by Vivek Khandelwal <[email protected]>
emitError is intended for error cases and not match failures of
patterns. notifyMatchFailure is...
Remove the linking of libtorch/ paths in delocate for CI builds
baa4383c44a066d4ac9036e93f667842e72adb10 authored over 2 years ago by powderluv <[email protected]>795479a88d8b5d5e244f59f94b4585b08c677dc0 authored over 2 years ago by Sean Silva <[email protected]>
d70bb68c9e8efd4fa940cddeb36cdfb48e87c909 authored over 2 years ago by Maksim Levental <[email protected]>
The biggest change here is to upgrade RefineTypes to the new sparse
dataflow framework.
Smaller...
85858d2743f174f9c967202f0d00c53ee384d575 authored over 2 years ago by Sean Silva <[email protected]>
The original conversion pattern for `AtenBatchNormOp` required that
the input rank be greater t...
This commit adds the decomposition for `aten.var.dim` op.
This commit also make changes in the d...
Remove all the libtorch downloads. If the user sets
-DTORCH_MLIR_USE_INSTALLED_PYTORCH=OFF then...
A previous patch added a new file
("VerifyConversionToValueSemantics.cpp") to the build, but it...
Found while trying to build torch-mlir on an AArch64 Linux VM, worth
a belts and braces to prev...
This patch adds a new pass `torch-verify-conversion-to-value-semantics`,
which looks for non-va...
A previous patch used lowercase names for LIT variables. This patch
replaces them with upperca...
Signed-off-by: Suraj Sudhir <[email protected]>
5e2012c7dd1f3e66e53465d53ca9d4c8d2b32fa7 authored over 2 years ago by Suraj Sudhir <[email protected]>lowering.
This commit addresses the remaining comments on lowering of
slice_scatter and select_s...
Prior to this patch, the canonicalizers for `AtenSizeOp` and
`AtenSizeIntOp` succeeded only if ...
The way we did it previously still created the slot and copied the
initializer even if unused.
Prior to this patch, the code in the `torch-simplify-shape-calculations`
pass iterated on the u...
1bb990afc7347bf5bf9432ef93bfda33a9f2ddd8 authored over 2 years ago by Maksim Levental <[email protected]>
This commit does three things:
1. Reverts some of the shape lib changes merged in
https://...
`aten.select_scatter` op.
This commit adds:
1. Lowering of `aten.slice_scatter` op into `tensor...
a08ff0d7f2c4388689428b0acccc4bf8d77ced14 authored over 2 years ago by George Petterson <[email protected]>
Set default OSX arch to x86_64. Release builds will override it.
Also update to the latest poin...
A lot of code was super verbose with `torch_mlir.OutputType.XYZ`. Now,
you can simply do `"xyz"`...
The canonicalizer converts `torch.prim.dtype` ops into integer constants
for valid types, but t...
There were some leftovers.
5bd9362c61828870cf3c4097c718ea34f475b11a authored over 2 years ago by Sean Silva <[email protected]>3ad810a1fb3648c764670b56a24e6f0fdb8dddb3 authored over 2 years ago by Henry Tu <[email protected]>
This commit adds support for lists of type `list<optional<tensor>>`
where each element in the l...
Temporarily revert to using PyTorch binaries until source builds
are ready to land.
TORCH_ML...
f202ae00129e5f03b14e3d04d63e3eca4da17ed5 authored over 2 years ago by powderluv <[email protected]>* Document workaround for PyTorch skew
* Update with review feedback
3d1a160101ed84ca26e6f87128360712173e73ca authored over 2 years ago by powderluv <[email protected]>
This patch replaces the use of raw integers like 6, 4, etc. (that
represent PyTorch's scalar ty...
Signed-off-by: Suraj Sudhir <[email protected]>
d38f2cae5bcf833f062241f0254a1d3d09ab2609 authored over 2 years ago by Suraj Sudhir <[email protected]>f0c3b5a7edd1084c108732a426724f53a1ee9001 authored over 2 years ago by Quinn Dawkins <[email protected]>
A prior patch (63538de2) that added support for bfloat16 type did not
add the canonicalization ...
Co-authored-by: Andrew Cain <[email protected]>
6885f1ed8a1b01a9f27ff5157c0a0500995c8151 authored over 2 years ago by Andrew Cain <[email protected]>
On my local machine, `unzip` didn't exist (producing a "command not
found" error), but CMake ig...
bbb648410ec35f335e7e807c07658d0a7ca0d8c6 authored over 2 years ago by Ramiro Leal-Cavazos <[email protected]>
Add an option to cache libtorch/ releases if you don't want to
download the latest. Add an opti...
Fixes https://github.com/llvm/torch-mlir/issues/994
be3d14cf76f53bc1a0d04ee04e1660b3f63af4cb authored over 2 years ago by powderluv <[email protected]>a1947c7bd12d341f9b9a62fc4f1756ac888c8213 authored over 2 years ago by powderluv <[email protected]>
* [MLIR][TORCH] Add decomposition of aten.repeat
* refine & rebase
* refine static shapes
...
A previous fix to the handling of size-1 dims in
`aten.view` (https://github.com/llvm/torch-mli...
TorchScript nodes like `prim::Load` and `prim::Store` aren't supported
in torch-mlir because th...
Enables BERT legalization.
Signed-off-by: Suraj Sudhir <[email protected]>
bb576c2cb372a865264c7ea33bf7584b7c4b4385 authored over 2 years ago by Suraj Sudhir <[email protected]>This moves torch-mlir to link against libtorch on macOS and linux
TESTS: Tests pass. Tested r...
2b52da951ba87f682b4eb629e791a263848a7b85 authored over 2 years ago by powderluv <[email protected]>
- Mention that Bazel build is in peripheral support tier
- Link to the e2e op wiki page
I ran into this while poking around at
https://github.com/llvm/torch-mlir/issues/959
cd79538a0c4ce626eb65adeaa83dc4de951a24b5 authored over 2 years ago by powderluv <[email protected]>
5fbf2a376c3d36f1f31a2cda8a242bd36f090383 authored over 2 years ago by Tanyo Kwok <[email protected]>
5888c4f7dc5aad504ef8c77442ebbf3c0393623c authored over 2 years ago by JakopinA <[email protected]>
In the `pyhpc_turbulent_kinetic_energy` TorchBench benchmark, the shape
calculation occurs insi...