Skip to content

fix: remove split.Tensor and split_with_sizes from decomp tables #4340

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
+1 −25
Open

fix: remove split.Tensor and split_with_sizes from decomp tables #4340

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
a8f4ea6
remove split and split_with_size from decomp tables
raayandhar Oct 10, 2025
ea723c1
add recompose to pass
raayandhar Oct 14, 2025
18142e6
remove tests from expected fail
raayandhar Oct 14, 2025
55c7704
remove tests from expected fail, hopefully correct this time
raayandhar Oct 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions lib/Dialect/Torch/Transforms/Passes.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,7 @@ void mlir::torch::Torch::createTorchDynamoExportToTorchBackendPipeline(
if (options.decompose) {
pm.addNestedPass<func::FuncOp>(
Torch::createDecomposeComplexOpsPass(options.backendLegalOps));
pm.addNestedPass<func::FuncOp>(Torch::createRecomposeComplexOpsPass());
pm.addNestedPass<func::FuncOp>(createCanonicalizerPass());
}
}
Expand Down
21 changes: 0 additions & 21 deletions projects/pt1/e2e_testing/xfail_sets.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -922,17 +922,6 @@
"UpSampleNearest2dVecNoneShape_basic",
"ViewCollapseDynamicWithAtenSizeIntModule_basic",
"ViewSizeFromOtherTensor_basic",
# Error: `aten.as_strided` op is not supported
"ChunkListUnpackDynamic_Module_basic",
"ChunkListUnpackUnevenDynamic_Module_basic",
"ChunkListUnpackUneven_Module_basic",
"ChunkListUnpack_Module_basic",
"SplitTensorGetItem_Module_basic",
"SplitTensorLastSmallerModule_basic",
"SplitTensorListUnpackModule_basic",
"SplitTensorNegativeDimModule_basic",
"SplitWithSizesListUnpackModule_basic",
"SplitWithSizes_Module_basic",
"Unfold_Module_basic",
"Unfold_Module_Rank_4",
"Unfold_Module_Rank_Zero_basic",
Expand Down Expand Up @@ -4012,17 +4001,7 @@
"AtenAsStridedModule_basic",
"AtenAsStridedNoStorageOffsetModule_basic",
"AtenAsStridedUnknownSizeModule_basic",
"ChunkListUnpackDynamic_Module_basic",
"ChunkListUnpackUnevenDynamic_Module_basic",
"ChunkListUnpackUneven_Module_basic",
"ChunkListUnpack_Module_basic",
"NativeGroupNormModule_basic",
"SplitTensorGetItem_Module_basic",
"SplitTensorLastSmallerModule_basic",
"SplitTensorListUnpackModule_basic",
"SplitTensorNegativeDimModule_basic",
"SplitWithSizesListUnpackModule_basic",
"SplitWithSizes_Module_basic",
# error: argument must be a memref of f32, f64, i32, i64, i8, i1, c32, c64, but got 'memref<3x5xbf16>'
"ElementwiseClampMaxModule_bfloat16",
"ElementwiseClampMinModule_bfloat16",
Expand Down
2 changes: 0 additions & 2 deletions projects/pt1/python/torch_mlir/dynamo.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,8 +52,6 @@ def _get_decomposition_table():
# support for aten.native_batch_norm_backward.
aten._native_batch_norm_legit_functional,
aten.native_group_norm,
aten.split.Tensor,
aten.split_with_sizes,
aten.norm.ScalarOpt_dim,
aten.embedding_dense_backward,
aten.native_layer_norm_backward,
Expand Down
2 changes: 0 additions & 2 deletions python/torch_mlir/extras/fx_decomp_util.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,8 +10,6 @@
torch.ops.aten.norm.ScalarOpt_dim,
torch.ops.aten.native_group_norm,
torch.ops.aten.upsample_bilinear2d.vec,
torch.ops.aten.split.Tensor,
torch.ops.aten.split_with_sizes,
torch.ops.aten.native_layer_norm,
torch.ops.aten.masked_fill.Tensor,
torch.ops.aten.masked_fill.Scalar,
Expand Down
Loading