Commit Graph

2859 Commits (6b95dd461d3e4a1b163de02f0ce998920b9f2500)
 

Author SHA1 Message Date
Vivek Khandelwal 6844c84702
[MLIR][Torch] Fix OnnxToLinalg lowering for AvgPool op (#3076)
Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
2024-04-01 22:14:14 +05:30
Stella Laurenzo 282e9b0e64
[fx] Fix type determination for multi-return ops and static `None` returns. (#3081)
In practice, this was caught by the way that AOT autograd traces
`convolution_backward`. For the unit test, we just repro it with a
custom op.
2024-04-01 09:39:38 -07:00
Gaurav Shukla 129a79417a
[MLIR][ONNX] Fix onnx.gather_nd implementation (#3070)
The indices should be expanded before the torch.gather operation.

Signed-off-by: Gaurav Shukla <gaurav@amd.com>
2024-04-01 20:17:09 +05:30
Xinyu Yang da88efad89
[Torch] Fix bug of DecomposeAtenSelectIntOp (#3087)
Fix bug of DecomposeAtenSelectIntOp. Because it may use resultTy when
resultTy has not been inferred.

```
    auto resultTy = op.getType().cast<BaseTensorType>();
    if (sliceTy.getSizes().size() == resultTy.getSizes().size()) {
      rewriter.replaceOp(op, slice);
      return success();
    }

```

So I add restriction.
2024-04-01 21:25:02 +08:00
Jiawei Wu 76080936d4
[stablehlo] add aten.index_put and aten.scatter_add op conversion support (#3086) 2024-04-01 19:39:49 +08:00
Stella Laurenzo 6d680ff445
[ods] Allow all tensor returns to be optional. (#3082)
This was found while tracing backwards graphs: the convolution_backwards
op will return None if the first result is not needed. Confirmed by
defining a custom op with a `Tensor` return signature and having its
meta kernel return None.
2024-03-29 23:09:34 -07:00
Xinyu Yang 40008b025a
[Torch] Support prelu decomposition (#3069) 2024-03-29 08:05:00 +08:00
zjgarvey c19fc9ba47
[ONNX] Fixes Issue with Dynamic Dims in GlobalAveragePool -> Torch Conversion (#3053)
Two e2e tests (AdaptiveAveragePool1/2dUnitOutputSizeDynamic) were
failing due to numerics. This was as a result of passing -1 as the
kernel size in the lowering for the corresponding onnx op
GlobalAveragePool.
2024-03-28 09:43:09 -07:00
Xinyu Yang e6e7689a24
[Torch] support decompose aten.einsum with ellipsis slicing (#3056) 2024-03-27 12:42:10 -07:00
Xida Ren (Cedar) 5f325749f9
add lowerings for AtenLtIntOp and AtenLeIntOp (#3061)
Co-authored-by: Xida Ren <xida.ren.dev@gmail.com>
2024-03-27 10:06:43 -07:00
Sambhav Jain 7825e12483
[Bazel] Add dep from `TorchPasses` to `TorchBackendTypeConversion` (#3065)
https://github.com/llvm/torch-mlir/pull/3055 adds
`lib/Dialect/Torch/Transforms/ScalarizeShapes.cpp`, which depends on
`torch-mlir/Dialect/TorchConversion/Transforms/BackendTypeConversion.h`.
```
ERROR: /root/.cache/bazel/_bazel_root/b89349c08f7224396763d14fe35cba11/external/torch-mlir/BUILD.bazel:170:11: Compiling lib/Dialect/Torch/Transforms/ScalarizeShapes.cpp failed: (Exit 1): clang failed: error executing command /usr/lib/llvm-16/bin/clang -U_FORTIFY_SOURCE -fstack-protector -Wall -Wthread-safety -Wself-assign -Wunused-but-set-parameter -Wno-free-nonheap-object -fcolor-diagnostics -fno-omit-frame-pointer ... (remaining 101 arguments skipped)

Use --sandbox_debug to see verbose messages from the sandbox and retain the sandbox build root for debugging
external/torch-mlir/lib/Dialect/Torch/Transforms/ScalarizeShapes.cpp:18:10: fatal error: 'torch-mlir/Dialect/TorchConversion/Transforms/BackendTypeConversion.h' file not found
#include "torch-mlir/Dialect/TorchConversion/Transforms/BackendTypeConversion.h"
         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.
Target @torch-mlir//:torch-mlir-opt failed to build
```

This PR adds the dependency and brings bazel builds back to green.

CI:
https://github.com/sjain-stanford/torch-mlir/actions/runs/8445558053/job/23132941876
2024-03-26 19:22:42 -07:00
Yuanqiang Liu 0a581a97a7
[Torch Dialect] enhance aten.int.tensor's canonicalize (#3058)
support fold with literal vtensor.  
change it to canonicalize because this pattern will create new op.
2024-03-27 09:51:58 +08:00
Stella Laurenzo e2343cf4ce
[fx] Implement auto_functionalized higher order op. (#3063)
* Also adds the basic scaffolding for handling more of these, which will
be needed for cond, while, etc.
* Refactors some of the support in the generic OpOverload emitter so it
can be shared with these other special forms.

This has been on my list for a while, but it just so happens that as
part of upgrading to PyTorch 2.3 and a pure upstream flow in Turbine, we
were using a feature that required integration with auto_functionalized.
This is perhaps the "weirdest" of the higher-order ops and a poor place
to start, but needs must. We have testing for this in Turbine.

Full support in Turbine has an entire custom ops facility. I've reduced
this down to a unit test in torch-mlir.
2024-03-26 17:06:05 -07:00
saienduri 11eaba3097
Create e2eshark_build.sh (#3062)
This commit creates a build script used by e2eshark test suite CI
2024-03-26 16:37:54 -07:00
Rob Suderman 14b548f968
[torch] Improve shape inference for `torch-to-linalg` path for reshapes (#3055)
Reshaping tensors depend on directly matching individual dimensions to
their corresponding dim in the `torch.view` reshape dimensions. This
involves decoupling dynamic dimensions from their static counterparts
and support cleanup / canonicalization.
2024-03-26 12:41:40 -07:00
Stella Laurenzo 17eeac880a
[fx] Accept `func_visibility=` and return created func op. (#3054)
This is a partial landing of #3046 while waiting for an upstream change
for the rest of it.
2024-03-25 16:48:06 -07:00
Vivek Khandelwal 9ae33e482e
[MLIR][TORCH] Add OnnxToTorch lowering for ops (#3049)
This commit adds the OnnxToTorch lowering for the Mish, Softplus,
HardSwish, Trilu, ThresholdedRelu op

Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
2024-03-25 20:29:07 +05:30
schnkmwt 1fcbfa87ec
Implement linalg lowering of diag_embed torch op (#2885)
This PR adds lowering of diag_embed to linalg dilect.
Tracked in https://github.com/nod-ai/SHARK-Turbine/issues/288

---------

Co-authored-by: sachink <sachink@xilinx.com>
2024-03-22 16:32:50 -07:00
zjgarvey 99b3a5f117
Converts all Adaptive Pooling Ops to Linalg (#2808)
The previous conversions for AtenAdaptiveAvgPool1dOp and
AtenAdaptiveMaxPool2dOp are refactored into a general templated
conversion that works for all of the AtenAdaptive...PoolNdOp's.

New support is added for the following ops:

1. AtenAdaptiveMaxPool1d
2. AtenAdaptiveMaxPool3d
3. AtenAdaptiveAvgPool3d

Support is also provided for passing inputs without batch dimensions.
For example, applying adaptive_avg_pool2d to an input tensor of rank 3.

After [pytorch #118162](https://github.com/pytorch/pytorch/pull/118162)
gets down to torch-mlir, I'll add a test for AdaptiveMaxPool1d with
return_indices (which will pass with that upstream fix).

---------

Co-authored-by: James Newling <james.newling@gmail.com>
2024-03-22 11:05:20 -07:00
zjgarvey 6aa481c204
[ONNX] LogSoftmax to Torch (#3024)
This PR adds support for onnx.LogSoftmax both for old versions (<13,
with axis >=0), and new versions (13).
2024-03-22 11:01:39 -07:00
Gaurav Shukla 50635dd509
[ONNX][MLIR] Add support for onnx.gather_nd (#2988)
Signed-off-by: Gaurav Shukla <gaurav@amd.com>
2024-03-22 21:38:39 +05:30
Stella Laurenzo 6ea857c644
[fx] Make the lift_fresh_copy -> clone special form use kwargs. (#3045)
At some point, this op became kwarg-only instead of arg/kwarg.
Discovered when upgrading to PyTorch 2.3.

Also adds a test as this was untested in-tree (was caught out of tree).
2024-03-21 15:34:40 -07:00
penguin_wwy 7616d637fd
Add stateless fx graph import (#3036) 2024-03-21 14:44:54 -07:00
Rob Suderman 3a56714bff
[torch] Fix clamp ranges on quantize_per_tensor on unsigned (#3018)
SExtValue was used for `int` and `uint` clamp values. This caused the
result to always be outputed as `zero`.
2024-03-20 13:37:47 -07:00
Xida Ren (Cedar) cb5cb506df
Fix SCF Forloop fails to convert to linalg when a tensor argument is supplied to the loop block (#3040)
Co-authored-by: Rob Suderman <rob.suderman@gmail.com>
Co-authored-by: Xida Ren <xida.ren.dev@gmail.com>
2024-03-20 11:04:02 -07:00
zjgarvey 6ff71b40c8
[ONNX] onnx.DynamicQuantizeLinear to Torch (#3009)
This adds support for converting DynamicQuantizeLinear from torch-onnx
to torch.

I could not get an e2e test to pass, since there seems to be some issues
with uint8 casting somewhere lower in the pipeline. For example
compiling with IREE for llvm-cpu, I would get either the correct zero
point (if zp < 128) or the correct zero-point minus 256 (if zp >= 128).
The output tensor seems to always return a tensor of zeros, which also
occurs when running uint8 examples through QuantizeLinear.

Edit: the first problem can be resolved by casting the output back to
uint8 on output, the second problem is resolved with PR #3018
2024-03-20 10:58:25 -07:00
jinchen 9cf6c45a39
Add OnnxToTorch support for Compress op (#3025) 2024-03-20 17:12:08 +00:00
Vivek Khandelwal 90e3d69c25
build: manually update PyTorch version (#3034)
Set PyTorch and TorchVision version to nightly release 2024-03-18.

Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
2024-03-20 21:45:07 +05:30
Aart Bik fe59f1ee0d
[torch-mlir][sparse] higher dimension COO (#3042)
Lift this from 2-dim only to n-dim for n>=2
2024-03-19 15:59:07 -07:00
Abhishek-TyRnT df02692726
Dynamic size support for flatten (#3005)
Added support for dynamic shapes in `flattenusingints` op in tosa
dialect. Due to this some Argmax tests pass
This PR fixes this issue https://github.com/llvm/torch-mlir/issues/3004

The following tests pass after this PR
 ```
1. "ArgmaxIntModule_basic"
2. "ArgmaxIntModule_multiple_maxs"
3. "ArgmaxModule_basic"
```
2024-03-19 15:19:29 -07:00
zjgarvey 7a9608bb69
[ONNX] Reduces onnx.Div sinceVersion to 7 (#3041)
The only difference between version 7 and newer versions is support for
different data types. We should allow this pattern to match as early as
7. Earlier versions have a more manual broadcast specification through
attributes, so I did not include those versions.

See: [onnx.Div
docs](https://onnx.ai/onnx/operators/onnx__Div.html#l-onnx-doc-divl)
2024-03-19 13:35:05 -07:00
Yuanqiang Liu 8b96727d0d
[Stablehlo] lowering chlo to stablehlo in torch-to-stablehlo pipeline (#3037)
as that stablehlo is better than chlo as the boundary between frontend
compiler and backend compiler.
2024-03-19 21:18:54 +08:00
Xida Ren (Cedar) 895ea8663a
add llvm style guide 2024-03-18 18:25:22 +00:00
Pavani Chowdary c51e2130f2
[onnx] support for lowering mod op from onnx to torch (#2859)
nod-ai/Shark-Turbine#267

---------

Authored-by: boddu.pavani@research.iiit.ac.in
Co-authored-by: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
2024-03-18 17:54:37 +05:30
Xinan Jiang(姜曦楠) d8a52e82c2
[onnx] Fix onnx.cast cases between int32 and int64 (#2982)
2 modifications:
1. torch.int64 is enum 4 in TORCH_DTYPE_TO_INT
2. add int32 support
2024-03-15 17:14:09 +00:00
penguin_wwy f34c187ac4
Normalize type hints to be compatible with multiple Python versions (#3028)
Although we provide a wheel package for Python 3.8, it may actually
throw the following exception:
`TypeError: 'type' object is not subscriptable`
2024-03-15 08:29:48 -07:00
Yuanqiang Liu 4282eb9e76
[Torch Dialect] support aten.fake_quantize_per_tensor_affine (#3014) 2024-03-15 08:53:29 +08:00
Nithin Meganathan 798bfd7dff
Adds accumulator types in TorchToLinalg for `AtenMmOp` and `AtenConvolutionOp` (#3027) 2024-03-14 16:40:40 -07:00
Sambhav Jain 0b2f9c89a2
Bring back `dynamic_shapes` constraints in fx importer API (#3026)
https://github.com/llvm/torch-mlir/pull/2992 dropped `constraints` from
the fx importer API,
[breaking](https://github.com/cruise-automation/mlir-tcp/actions/runs/8284385380/job/22669774071)
downstream AOT compile tests in `mlir-tcp` that use it. This knob has
been soft-deprecated for a while now, replaced by `dynamic_shapes` - a
more ergonomic interface. This PR brings back dynamic_shapes constraints
in the new supported form. Also added a python lit test with dynamic
shaped annotations.
2024-03-14 10:26:34 -07:00
penguin_wwy 29ac23a790
Setuptools uses a separate build directory (#3023)
* setuptools not steal the build directory name
https://github.com/llvm/torch-mlir/pull/3021#issuecomment-1994447855
* support pre-built LLVM
* support CMAKE_BUILD_TYPE env
2024-03-13 20:41:48 -07:00
Yuanqiang Liu 870e63bc3c
[Torch Dialect] support decomposition of aten.linspace (#3006) 2024-03-14 08:28:33 +08:00
Yuanqiang Liu 43c6996a31
[Torch Dialect] add folder for aten.ceil and unify patterns of ceil, … (#3010)
…floor, round
2024-03-14 07:41:58 +08:00
ptrifunovic98 524ff99216
Implement lowering of torch.aten.linalg_cross (#2986)
Closes
[nod-ai/SHARK-Turbine#497](https://github.com/nod-ai/SHARK-Turbine/issues/497)
2024-03-13 12:17:22 -07:00
aldesilv 6fa21bd8b1
OnnxToTorch lower celu op (#2920) 2024-03-13 20:34:10 +05:30
Nithin Meganathan 5ecc1d5c0d
Align softmax accumulation types with Torch's CUDA implementation (#2996) 2024-03-12 15:07:45 -07:00
Yuanqiang Liu ad6159c7cb
[Stablehlo] lowering aten.round to stablehlo.round_nearest_even (#3011) 2024-03-12 08:58:20 +08:00
Rob Suderman e78c99e74e
[torch] Update folders for splat operators (#3012)
Splat operators required the output is 1-D. This was not a required
restriction and was loosened to 2d.
2024-03-11 16:45:49 -04:00
Devjiu 4b1e87ce67
[TorchDynamo] Enable Elemtwise ops for Scalar arg (#2744)
This commit provides dummy solution to support elmentwise operations
(mul, add) with scalar argument. ( op(Tensor, Scalar) )

It replaces `torch.aten.add.Tensor` with `torch.aten.add.Scalar`.
```
Unexpected outcome summary: (torchdynamo)

****** Unexpectedly Passed tests - 22 tests
    XPASS - "AddCDivModule_basic"
    XPASS - "BatchNorm1DModule_basic"
    XPASS - "BatchNorm1DStaticShapeModule_basic"
    XPASS - "BatchNorm1DWith2DInputModule_basic"
    XPASS - "BatchNorm2DModule_basic"
    XPASS - "BatchNorm3DModule_basic"
    XPASS - "ElementwiseAddScalarInt64Module_basic"
    XPASS - "ElementwiseAddScalarIntModule_basic"
    XPASS - "ElementwiseMulScalarModule_basic"
    XPASS - "ElementwiseMulScalarModule_float"
    XPASS - "ElementwiseMulScalarModule_int"
    XPASS - "GroupNormModule_basic"
    XPASS - "GroupNormNoWeightAndBiasModule_basic"
    XPASS - "MobilenetV3Module_basic"
    XPASS - "NativeBatchNorm1DModule_basic"
    XPASS - "NativeBatchNorm2DModule_basic"
    XPASS - "NativeBatchNorm3DModule_basic"
    XPASS - "NativeBatchNormNoneWeightModule_basic"
    XPASS - "NativeGroupNormBackwardModule_basic"
    XPASS - "NativeGroupNormModule_basic"
    XPASS - "ResNet18Module_basic"
    XPASS - "ResNet18StaticModule_basic"
```

And segfault for test
"ElementwiseAddScalar_TensorLiteralInt32_Module_basic". Somehow this
change doesn't allow to use Tensors, that are not forward arguments, but
local variables of model.
e.g. `self.x = torch.tensor(..)`

See also: #2745

Signed-off-by: Dmitrii Makarenko <dmitrii.makarenko@intel.com>
2024-03-11 12:22:05 -07:00
Rob Suderman 8fb28661f9
[onnx] Fix onnx.ReduceMean lowering (#3002)
Reduce mean lowerings did not succesfully lower to `linalg` via torched.
There were two separate paths that could be consolidated to a single
simpler pass. This resulted in a significant improvement in test
coverage.
2024-03-11 11:32:53 -07:00
Yuanqiang Liu 229ca3a9e1
[Torch Dialect] emit aten::mul and add folder (#3007) 2024-03-11 19:59:34 +08:00