The added e2e maxpool testcase from #545 was not getting a static shape
due to an unfolded prim....

* external/llvm-project 881ff4e4ebe8...84fe34a0b7fd (466):
> [MLIR] Workaround for python det...

* [tosa] Support for AtenBatchNormOp

Signed-off-by: Anup Gangwar <[email protected]>

Co...

This commit adds a test module specifically for table batch embedding
algorithm. This test case ...

Signed-off-by: Suraj Sudhir <[email protected]>

Refactor matmul into separate class and derive variants:
- matmul
- mm, bmm
- linear

Signe...

* [tosa] Support for AtenCeilOp and AtenReciprocalOp
* [tosa] Support for comparator ops, Aten[...

In addition, use add_llvm_executable so that the relative rpath of
the installed binaries is se...

Signed-off-by: Suraj Sudhir <[email protected]>

This commit adds lowering of `aten.index.Tensor` op

Signed-Off By: Vivek Khandelwal <vivek@nod-...

- Also handles braodcasting n-D tensors, dynamic shapes

Signed-off-by: Suraj Sudhir <suraj.su...

This reverts commit c9a343267c863186d5127fb8e38a76f62a53e2ac.

- Add `qualified` to ods because of
https://reviews.llvm.org/D113873 and https://reviews.llvm.or...

- Common code as TF repository, being moved to MLIR core.
- Will support further legalizations ...

Signed-off-by: Suraj Sudhir <[email protected]>

This commit adds the aten ops which do not require random number
support to aten dialect. This c...

disable ElementwisePowModule_basic
* Update nll_loss_forward to align to the change in PyTorch
...

Signed-off-by: Anup Gangwar <[email protected]>

Note that to enable folding of the code coming from an example
like the ConstantPad2dStaticModul...

This commit adds support for two return values of type
memref f32 and i64.

Signed-Off By: Vivek ...

This commit adds lowering of `aten.threshold` op
This commit adds lowering of `aten.threshold_bac...

The element type of tensor literal should be signless when converted to
builtin tensor types.

Signed-off-by: Suraj Sudhir <[email protected]>

This involes the following 2 parts:
- Change refine type to propagate more static shape info.
- ...

Templated variants for squeeze and squeeze.dim

- This commit adds E2E support for `aten.ones_like` and
`aten.zeros_like` ops.
- Adds support ...

This commit adds include statements of the file
`torch/csrc/jit/ir/ir.h` for files that use type...

This commit adds lowering of `aten.arange.start_step` op.
This commit decomposes `aten.arange` an...

- It folds `aten.to.dtype` when the input tensor type and result type
are exactly same.
- It f...

`aten.add.Scalar` and `aten.sub.Scalar` op conversions have been added.
The changes have been ma...

We only handle the expanding OR collapsing cases, we do not handle
expanding And collapsing happ...

This commit adds decomposition of `aten.empty_like` into `aten.empty`
op.

Signed-Off-by: Gaurav...

- Added E2E support for `aten.eq.Tensor` and `aten.lt.Tensor` ops. Both
the operands are expec...

Signed-off-by: Anup Gangwar <[email protected]>

This commit fixes the naming of results in the torch ODS generator
when dealing with multiple re...

Signed-off-by: Suraj Sudhir <[email protected]>

The existing implementation of `ConvertConstantTensorAllocOp<>` requires
a C++17 feature `if con...

Add the required lowerings and correct test cases.
These op produce zero-d tensors and it was in...

Signed-off-by: Anup Gangwar <[email protected]>

Co-authored-by: Anup Gangwar <anup.gangwar...

Support for integer types in gtScalar op has been added.
The code share same logic with gtTensor...

- Templatize `aten.zeros` and `aten.ones` ops lowering.
- Add E2E support for `aten.empty` op.
-...

`aten.gt.Tensor` op has been added in torch dialect and the
lowering of the op has been done to ...

This commit adds lowering of `aten.ceil` op as a part of element-wise
ops lowering.

Signed-Off-...

This patch adds the where, gt, bucketize and reshape
ops to the Torch dialect. These ops are pre...

This commit adds support for aten.native_layer_norm operation. Here
the previous code for aten....

This commit adds lowering of `aten.squeeze.dim` op into
`linalg.TensorCollapseShape` op. Here, t...

This commit adds lowering of `aten.index_select` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-...

https://github.com/llvm/torch-mlir/pull/464#discussion_r765065092

Signed-Off By: Vivek Khandelw...

This commit adds lowering of `aten.gt.Scalar` and `aten.where.self` as a
part of element-wise op...

Previously aten.conv2d was only lowered if there was no bias.
Here lowering is extended to suppo...

Support for passing memref of bool types as a function argument
and return is added in ref-backe...

This commit adds lowering of `aten.zeros` op.

Signed-Off By: Vivek Khandelwal <[email protected]>

The op lowering has been added as a part of `torch-lower-to-linalg`
pass. This takes care of ign...

The external llvm is updated to point to
https://reviews.llvm.org/rG966b72098363d44adf2882b9c34f...

- Supports variants with multiple dims, one dim, all dime
- Leverages legalize_common and legali...

There is an op name change that requires trivial changes.
Also, some of the warning has been fix...

This commit adds lowering of `aten.bitwise_and.tensor` op.

Signed-Off By: Vivek Khandelwal vivek...

This commit adds lowering of `aten.mean` and `aten.numel` op.

Signed-Off By: Vivek Khandelwal <...

Fixes broken e2e test ElementwiseAddModule_basic

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm....

Many reduction ops take as an argument an optional output dtype that
can change the type of the ...

This commit adds lowering of `aten.Squeeze` op into
`linalg.TensorCollapseShape` op. The size 1 ...

This is to fold the common pattern from Bert inference like:
```
%111 = torch.prim.NumToTensor.S...

The lowering of aten.fill.Scalar has been added.
The changes have been made as a part of -torch-...

Co-authored-by: dan <[email protected]>

Signed-Off-By: Prateek Gupta <[email protected]>

This commit adds lowering of `aten.div.Scalar`.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs....

This commit fixes a type promotion bug when NumToTensor was given a
float as an argument. In par...

Share code in `log_softmax_backward` and `softmax_backward` ops.

aten.log_softmax_back_data op lowering and required
tests has been added. Some NFC have also bee...

This commit adds lowering of `aten.mul.Scalar` and also adds
decomposition of `aten.addmm` to `a...

Now, aten::linear supports rank 3 inputs. This is a fix
for upcoming bert-inference task. The co...

This commit adds minor non functional changes to the aten.gelu_backward
operation.

Signed-Off-B...

This commit adds new operation `aten.gelu_backward` in the aten
dialect and adds lowering of th...

This change is to unblock the work of some backprop ops returning more
than one tensors. We will...

Update AsmPrinter/Parser and MatchAndRewrite

This commit fixes a bug with the way ConvertAtenOnesOp was matching on
the pin_memory bool argum...

- Remove use of conversion construction macros
- Add mul and div op conversions
- Add correspo...

Add a bunch of op legalizations.

Signed-off-by: Suraj Sudhir <[email protected]>

The e2e testing for aten_tanh_backward op has been added.
The testing is done for ref_backend.