llvm
/
torch-mlir
mirror of https://github.com/llvm/torch-mlir
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
3,105 Commits
47 Branches
1195 Tags
90 MiB
Commit Graph

93 Commits (f9766c89f6d055edc89563365c86efb5f892cc89)

Author SHA1 Message Date
Aart Bik f72770a725
[torch-mlir][sparse] replace ad-hoc mechanism with proper FX export (#3648) 
Now that the PyDev feature request pytorch/pytorch#117188 has been
completed, we can remove all the ad-hoc code that propagates sparsity
metadata and replace it with the built-int PyDev metadata for sparse
tensors. This removes a lot of code and also ensures sparsity is
consistent with the torch.sparse package for all cases.
 2024-08-20 09:56:21 -07:00
Matthias Gehre 334633b738
e2e: Enable generate-runtime-verification pass (#3615) 
This adds the `generate-runtime-verification` pass into the linalg
refbackend, and moves all tests that now abort at runtime into the crash
set, sorted by their respective errors.

I have fixed on set of errors found that way, which are mismatches
between the static dimensions we cast to and the actual dynamic
dimensions. This was caused by wrong annotations on the test cases, like
in
https://github.com/llvm/torch-mlir/pull/3615/files#diff-48bfbf41fcad5fa01b49197d251114f84a2b8de4f1d87ab938a061aedd1419b1R1931
 2024-08-12 14:15:12 +02:00
Vivek Khandelwal 22c9008bb9
build: Update Roll PyTorch version (#3548) 
This commit also updates the PyTorch and Torchvision nightly links since
they are now moved to a different location.

PyTorch Nightly: https://download.pytorch.org/whl/nightly/cpu/torch/
Torchvision Nightly:
https://download.pytorch.org/whl/nightly/cpu/torchvision/

Disables dtype checks for some ops, tracked by https://github.com/llvm/torch-mlir/issues/3552

Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
 2024-07-19 21:38:57 +05:30
Matthew Francis-Landau fe9db78120
Allow custom ops to return an array of tensors (#3531) 
This PR adds support to `fx_importer.py` for handling custom ops that
return an array of tensors. As long as the length of the array is
consistent across runs (determined statically), then this patch will
work. This does not require that the number of tensors returned is
determined by the op's definition.

CC @sjain-stanford
 2024-07-14 11:54:23 -07:00
Sambhav Jain 7411ff2f69
[Symbolic Shapes] Test coverage for unbacked symint from data dependent ops (#3542) 
We do have support for translating unbacked symbolic_ints that arise
from data-dependent ops like `aten.nonzero`. This PR adds the python lit
test coverage for the same.
 2024-07-14 11:52:03 -07:00
Aart Bik 6fece25ff3
[torch-mlir][sparse] add decomposition features to sparse compiler (#3505) 
Fixes https://github.com/llvm/torch-mlir/issues/3499
 2024-06-28 10:18:36 -07:00
Andrea 🦈 51902ec2dc
Create MLIR functions for ONNX operators that are functions (#3409) 
Resolves #3384.

Many ONNX operators are defined by functions and therefore could be
expanded into simpler ONNX operations during importing, avoiding the
need for tools downstream to support these operators directly.

This commit adds this capability to onnx_importer.py. When importing a
node, the schema for the node's operator is retrieved. If the schema
provides a function for the operator, a specialized version for the
node's types and attributes will be created and imported as an MLIR
function with private visibility. An MLIR function call will then be
emitted, instead of a normal operator node. Caching is used to avoid
generating redundant functions within the same module.

In order to avoid a disruptive change to the importer output for a
large number of operators that already have TorchOnnxToTorch support,
an allowlist strategy is used by default. With this commit, only one
operator is allowlisted for expansion, MeanVarianceNormalization.
However, many other operators can be correctly expanded by the current
code, so hopefully the allowlist can be gradually extended. It is
possible to disable the allowlist in the configuration, in which case
all functions are expanded (useful for testing).

Tools downstream of the importer may now need to do inlining when
consuming the output of the importer, e.g.:

  cat imported.mlir | torch-mlir-opt --inline --convert-onnx-to-torch

Explanations for subtle code changes:

- Looking up the correct schema and function for an operator requires
  knowing the opset version. NodeImporter retrieves this from the
  opset imports on the ModelProto retained by the GraphInfo. Previously,
  the model_proto field on GraphInfo was None when importing a subgraph
  in import_regions, but this conflicts with the new need for opset
  version info. Since the apparent purpose of setting it to None was to
  control how GraphInfo generates its input map, a new flag is added to
  GraphInfo (is_subgraph) to control this behavior, so that the actual
  ModelProto can now be provided without breaking this. This also turned
  out to be useful for getting the Config via ModelInfo via GraphInfo.
- Some operators' functions are context-dependent, which means the
  function definition depends on the types of the inputs. Therefore node
  importing now needs to look up the types of a node's inputs, not just
  its outputs as was the case previously. Consequently the operand to
  find_type_proto_for_name() may now be a graph input or initializer in
  some cases, so it has to be updated.
 2024-06-14 10:11:26 -07:00
Aart Bik d77bab37d1
[torch-mlir][sparse] re-enable all sparse tests (#3444) 
this fixes the following issue:

https://github.com/llvm/torch-mlir/issues/3418
 2024-06-10 11:19:32 -07:00
Sambhav Jain 7e0e23c668
Test custom op import with symbolic shapes (#3431) 
Tests the basic constructs of registering a custom op and its abstract
implementations (with FakeTensors) in python, going through TorchDynamo
export, followed by importing the shape expressions in the Torch
dialect.

Also fixes the importer were previously the symbolic bind op insertion
was not gated in one place.
 2024-06-09 00:32:49 -07:00
Sambhav Jain d0a818a03e
Representing Symbolic Shape Expressions in Torch Dialect (#3372) 
Torch Dialect with symbolic shape expressions:
```ll
module {                                                                                                                                                                                                     
  func.func @main(%arg0: !torch.vtensor<[?,?,3],f32>, %arg1: !torch.vtensor<[?,?,3],f32>) -> !torch.vtensor<[?,?,3],f32> {                                                                                   
    %0 = torch.symbolic_int "s0" {min_val = 5, max_val = 10} : !torch.int                                                                                                                                    
    %1 = torch.symbolic_int "s1" {min_val = 0, max_val = 100} : !torch.int                                                                                                                                   
    %2 = torch.symbolic_int "s3" {min_val = 0, max_val = 50} : !torch.int                                                                                                                                    
    
    torch.bind_symbolic_shape %arg0, [%0, %1], #affine_map<()[s0, s1] -> (s0, s1, 3)> : !torch.vtensor<[?,?,3],f32>                                                                                          
    torch.bind_symbolic_shape %arg1, [%0, %2], #affine_map<()[s0, s1] -> (s0, s1, 3)> : !torch.vtensor<[?,?,3],f32>                                                                                          
    
    %3 = torch.aten.tanh %arg0 : !torch.vtensor<[?,?,3],f32> -> !torch.vtensor<[?,?,3],f32>                                                                                                                  
    torch.bind_symbolic_shape %3, [%0, %1], #affine_map<()[s0, s1] -> (s0, s1, 3)> : !torch.vtensor<[?,?,3],f32>                                                                                             
    
    %4 = torch.aten.sigmoid %arg1 : !torch.vtensor<[?,?,3],f32> -> !torch.vtensor<[?,?,3],f32>                                                                                                               
    torch.bind_symbolic_shape %4, [%0, %2], #affine_map<()[s0, s1] -> (s0, s1, 3)> : !torch.vtensor<[?,?,3],f32>                                                                                             
    
    %5 = torch.prim.ListConstruct %3, %3, %4 : (!torch.vtensor<[?,?,3],f32>, !torch.vtensor<[?,?,3],f32>, !torch.vtensor<[?,?,3],f32>) -> !torch.list<vtensor>                                               
    %int1 = torch.constant.int 1                                                                                                                                                                             
    %6 = torch.aten.cat %5, %int1 : !torch.list<vtensor>, !torch.int -> !torch.vtensor<[?,?,3],f32>                                                                                                          
    torch.bind_symbolic_shape %6, [%0, %1, %2], #affine_map<()[s0, s1, s2] -> (s0, s1 * 2 + s2, 3)> : !torch.vtensor<[?,?,3],f32>                                                                            
    
    return %6 : !torch.vtensor<[?,?,3],f32>                                                                                                                                                                  
  }                                                                                                                                                                                                          
}              
```

For reference, this is the TorchDynamo exported program with symbolic
shape expressions that the above Torch dialect program is imported from:
```py
ExportedProgram:                                                                                                                                                                                             
    class GraphModule(torch.nn.Module):                                                                                                                                                                      
        def forward(self, x: "f32[s0, s1, 3]", y: "f32[s0, s3, 3]"):                                                                                                                                         
            # File: /home/sambhav.jain/workspaces/cruise/src/3p/torch-mlir/test/python/fx_importer/symbolic_shape_expr_test.py:31 in forward, code: a = torch.tanh(x)                                        
            tanh: "f32[s0, s1, 3]" = torch.ops.aten.tanh.default(x);  x = None                                                                                                                               
                                                                                                                                                                                                             
            # File: /home/sambhav.jain/workspaces/cruise/src/3p/torch-mlir/test/python/fx_importer/symbolic_shape_expr_test.py:32 in forward, code: b = torch.sigmoid(y)                                     
            sigmoid: "f32[s0, s3, 3]" = torch.ops.aten.sigmoid.default(y);  y = None                                                                                                                         
                                                                                                                                                                                                             
            # File: /home/sambhav.jain/workspaces/cruise/src/3p/torch-mlir/test/python/fx_importer/symbolic_shape_expr_test.py:33 in forward, code: return torch.cat((a, a, b), dim=1)                       
            cat: "f32[s0, 2*s1 + s3, 3]" = torch.ops.aten.cat.default([tanh, tanh, sigmoid], 1);  tanh = sigmoid = None                                                                                      
            return (cat,)                                                                                                                                                                                    
                                                                                                                                                                                                             
Graph signature: ExportGraphSignature(input_specs=[InputSpec(kind=<InputKind.USER_INPUT: 1>, arg=TensorArgument(name='x'), target=None, persistent=None), InputSpec(kind=<InputKind.USER_INPUT: 1>, arg=TensorArgument(name='y'), target=None, persistent=None)], output_specs=[OutputSpec(kind=<OutputKind.USER_OUTPUT: 1>, arg=TensorArgument(name='cat'), target=None)])                                               
Range constraints: {s0: ValueRanges(lower=5, upper=10, is_bool=False), s1: ValueRanges(lower=0, upper=100, is_bool=False), s3: ValueRanges(lower=0, upper=50, is_bool=False)} 
```

Huge credit to @stellaraccident for the inputs that helped evaluate the
various design options and arrive at the representation of choice.


- [x] Op definitions for symbolic_int and bind_symbolic_shape ops
- [x] fx_importer updates to import range constraints + create
symbolic_int ops
- [x] fx_importer changes for AffineMapAttr building + adding
bind_symbolic_shape ops
- [x] custom printer/parser for inlined AffineMap expressions in mlir
assembly
- [x] Dialect lit test
- [x] fx_importer python lit tests
- [ ] Cleanup pass to remove these ops (can add in a follow-on)
 2024-06-07 04:04:03 -07:00
Vivek Khandelwal 72837fbb3d
build: manually update PyTorch version (#3340) 
Set PyTorch and TorchVision version to nightly release 2024-05-14.

Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
 2024-06-06 22:23:40 +05:30
Aart Bik 560ca24771
[torch-mlir][sparse] replace xavier with ones initialization (#3374) 
ensures stability of results between different set ups
 2024-05-21 17:12:55 -07:00
Aart Bik c0e7d2667d
[torch-mlir][sparse] inference mode for sparse GCN test (#3369) 2024-05-20 19:52:16 -07:00
Aart Bik e80f072ba4
[torch-mlir][sparse] example of a sparse graph convolution (#3363) 2024-05-17 15:43:50 -07:00
Peiming Liu ccb772cd0f
[sparse] propagate sparsity properly when decompose torch operations. (#3318) 2024-05-15 10:09:27 -07:00
Aart Bik 44fa6c3afd
[torch-mlir][sparse] sparse diagonal feature scaling test (#3344) 2024-05-14 12:13:54 -07:00
Peiming Liu 8e74d64e8f
[sparse] convert to sparse before any use in sparse test. (#3337) 2024-05-14 09:10:36 -07:00
Aart Bik 667dfcbc5a
[torch-mlir][sparse] enable test on ReLu (#3336) 
Downstream MLIR sparsifier has some (rudimentary) support for ReLU now,
and this test can now be enabled with correct end-to-end behavior.

Also see discussion at:

https://discourse.llvm.org/t/min-max-abs-relu-recognition-starter-project/78918
 2024-05-13 15:34:26 -07:00
Peiming Liu 2c22087cab
[sparse] match fx node using target name instead of variables name (#3315) 2024-05-09 12:34:14 -07:00
Aart Bik 97a822de0a
[torch-mlir][sparse] minor tweaks in sparse tests (#3311) 
(1) test full pytorch output for eltwise
(2) use "random" input for LIF, to get general sparse tensor 
(3) introduce way to get true sparsity into network (needs backend fix
first)
 2024-05-09 10:03:25 -07:00
Aart Bik 89bb7404c1
[torch-mlir][sparse] add a true network to our NN tests (#3305) 
Objective: make the to_sparse work end-to-end!
 2024-05-08 21:18:42 -07:00
Peiming Liu cff144b3ac
[sparse] fix double free due to incompatibility between buffer-deallo… (#3303) 
…cation and sparse tensors.

**NOTE**: This PR _doges_ the issue in buffer-deallocation pass instead
of resolving it. In the future, we need to fix the bug in
buffer-deallocation pass when handling code generated by sparse
compiler.
 2024-05-08 21:18:17 -07:00
Aart Bik c4b28e8d9f
[torch-mlir][sparse] test for sparse "activation" (#3304) 
Example of introducing sparsity into the forward pass. With a bespoke
propagation (but upstream PyTorch will support this).
 2024-05-08 19:01:24 -07:00
Aart Bik c77f3b559a
[torch-mlir][sparse] add simple sparsity "propagation" rules (#3297) 
While waiting for the full resolution of feature request
https://github.com/pytorch/pytorch/issues/117188
(which will propagate sparsity the right way in upstream PyTorch for all
FX Graphs), this minor change allows us to start testing sparsity
"within" a network, rather than just the parameters. Feel free to add
your own rules for testing (but within reason for what will be done
upstream).

Note, two TODOs need to be addressed to work around some pending issues
to make the JIT execution work.
 2024-05-07 15:27:36 -07:00
Sambhav Jain 2176176fef
[FX] Add broadcast test with dynamic dim (#3123) 
This scenario was uncovered in a downstream test that failed with a
previous snapshot of torch-mlir. See
https://github.com/cruise-automation/mlir-tcp/actions/runs/8605480116/job/23581829102?pr=65.
```
  File "/home/runner/.cache/bazel/_bazel_runner/ce288f117ee4ca92dc028a6a28476a3d/sandbox/processwrapper-sandbox/2380/execroot/mlir-tcp/bazel-out/k8-opt-exec-2B5CBBC6/bin/test/AotCompile/broadcast_unit_dim_to_dynamic_with_unchanged_dim_dynamic_torch_exporter.runfiles/pip_deps_torch_mlir/site-packages/torch_mlir/extras/fx_importer.py", line 969, in value_info_to_type
    raise NotImplementedError(
NotImplementedError: Could not deduce type from value info: tensor_meta=None, val=s1, sparsity=None
```
It seems to have resolved on current HEAD. Adding this test to ensure
coverage in the future.
 2024-04-29 09:21:12 -07:00
Stella Laurenzo 6877302504
[NFC reformat] Applies pre-commit formatting to Python files. (#3244) 
This is a large change because prior to this point, Python files in the
project were not consistently formatted. This reformats them all with
black defaults.

Based on experience with prior projects, if you have a dev/long-term
branch with Python patches, you can minimize merge conflicts prior to
rebasing to include this commit by running `black` on your modified
Python files, squashing, and then rebasing/merging.
 2024-04-27 14:16:31 -07:00
Stella Laurenzo 5d4b803914 [NFC reformat] Run pre-commit on all files and format misc. 
This is part 1 of ~3, formatting all miscellaneous text files and CPP files matched by a first run of pre-commit. These tend to be low change-traffic and are likely not disruptive.

Subsequent patches will format Python files and remaining CPP files.
 2024-04-27 14:08:09 -07:00
Aart Bik 491f4820f5
[torch-mlir][sparse] pre-pend named buffers to parameter list (#3178) 
weights and biases and other model parameters appear as a separate data
structure to the traced graph, but are needed when running the MLIR
compiled code; this PR implements that extended functionality
 2024-04-17 14:44:05 -07:00
penguin_wwy af5509c5d9
[FxImporter] Type conversion to resolve the mismatch between Py type and schema type (#3163) 2024-04-15 23:14:19 -07:00
Aart Bik 307f49f566
[torch-mlir][sparse] support sparse tensor output (#3152) 
Sparse inputs and outputs are now fully supported! They always consist
of their constituents buffers, passed as numpy arrays. Sparse on!
 2024-04-12 09:56:32 -07:00
Aart Bik 184d8c13f4
[torch-mlir][sparse] add ID-net example (#3127) 
first sparse-in/sparse-out example, will be used
to make actual sparse output work!
 2024-04-09 11:21:30 -07:00
Aart Bik 5797d3aa57
[torch-mlir][sparse] add a COO test for 3-dim (#3119) 
This tests COO for more than 2-dim. Note that sparsity should really
propagate into the relu activation and the output, but such cleverness
needs to wait for the pending work in the PyTorch tree.
 2024-04-08 16:46:51 -07:00
Stella Laurenzo ffaaf08c31
[fx] Fix type inference for scalar/int types. (#3099) 
This was discovered in a downstream test suite and was due to a control
flow nesting merge issue. In-tree test added and fixed.
 2024-04-02 13:56:43 -07:00
Rob Suderman ec4cb8be44
Bump LLVM to llvm/llvm-project@0030fc4ac7 (#3079) 
Co-authored-by: Peiming Liu <peiming@google.com>
 2024-04-01 16:34:59 -07:00
Stella Laurenzo 826786bdd0
[fx] Support ExportedProgram buffer mutation. (#3080) 
In the prior state when I supported mutation of user inputs by treating
them as mutable-tensor SSA values, I had left the case of buffer
mutation only vaguely implemented until a concrete use emerged.
    
This patch reworks this buffer mutation support by assuming that buffers
must be resolved via the hooks symbolically and treated with load/store
semantics. This is implied in the structure since we have no SSA value
that represents a buffer and we already assume that reading parameters
happens via such a mechanism.
 2024-04-01 14:18:12 -07:00
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
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
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
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
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
Vivek Khandelwal 6e84752c39
build: manually update PyTorch version (#2992) 
Set PyTorch and TorchVision version to nightly release 2024-03-07.
This commit also removes the deprecated constraints API:
342e7929b8

Signed-Off By: Vivek Khandelwal <vivekkhandelwal1424@gmail.com>
 2024-03-07 21:42:38 +05:30
Aart Bik f21b76b68a
[torch-mlir][sparse] fixed merge conflict (#2967) 2024-02-28 17:14:00 -08:00
Peiming Liu e85a2a87c5
[torch-mlir][sparse] support e2e sparse kernels with COO inputs. (#2939) 2024-02-28 16:08:37 -08:00
Rob Suderman e48fe45886
[onnx] Import `onnx` import to pass remaining tests (#2951) 
Finish supporting importing the vast majority of `onnx` operations. This
includes:
- region support
- region value inherentance
- `torch.string` support
- `torch.list` support
- `torch.optional` support
 2024-02-28 12:18:02 -08:00
Aart Bik 30212547a9
[torch-mlir][sparse] add JIT test for block sparse SpMV (#2955) 
This required adding a "decompose" pass to the torch lowering, since
torch.mv was not directly handled by lowering to linalg
 2024-02-27 11:49:32 -08:00
Sambhav Jain 3cbe6c98ec
Expose `func_name` to the main fx import API (#2949) 
As titled.
 2024-02-26 10:08:14 -08:00
Aart Bik 4147b280ce
[torch-mlir][sparse] add block sparsity to mlir lowering (#2942) 
Also note that we are in the process of proposing SparseTensorMetadata
to PyTorch FX graph export (see
https://github.com/pytorch/pytorch/pull/117907). This will hopefully
eventually replace the current data structures in torch-mlir.
 2024-02-23 11:57:20 -08:00
Rob Suderman 53f6d06ab8
[onnx] Drop `ConstantOfShape` logic form importer, fix torch lowering (#2930) 
There is no reason to treat `ConstantOfShape` as a specialized import
any as there exists a onnx-to-torch equivalent. Dropping the import
coding and adding support for resource conversion substantially
increases test coverage for dynamically shaped tests.
 2024-02-21 21:34:43 -08:00
Rob Suderman 13553d49c9
[onnx] Update the importer to create a `none` for missing operands (#2931) 
Some operands are optional so we require a placeholder for missing
operands. We invent an `onnx.None` operation as our placeholder.
 2024-02-20 09:30:30 -08:00