Skip to content
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

build: manually update PyTorch version #3727

Merged
merged 15 commits into from
Oct 18, 2024
Merged

build: manually update PyTorch version #3727

Show file tree
Hide file tree
Changes from 14 commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
1eea6fe
build: manually update PyTorch version
vivekkhandelwal1 Oct 7, 2024
2d4d87c
Update filechecks for failing test
vivekkhandelwal1 Oct 7, 2024
b983142
Update torch version
vivekkhandelwal1 Oct 9, 2024
b5ffb20
This commit disables the failing sparse tensor tests since they are n…
vivekkhandelwal1 Oct 9, 2024
a7e171f
Fix lint checks
vivekkhandelwal1 Oct 9, 2024
a8f8afa
Run test for specific torch version
vivekkhandelwal1 Oct 10, 2024
aad34f8
Fix lint check
vivekkhandelwal1 Oct 10, 2024
8d50d50
Update PyTorch version
vivekkhandelwal1 Oct 15, 2024
4dd4cbd
Disable some filechecks due to different IRs for stable and nightly v…
vivekkhandelwal1 Oct 15, 2024
bdfacf0
Update xfail set
vivekkhandelwal1 Oct 15, 2024
128b7e3
Update xfail set for Fx_Importer tests
vivekkhandelwal1 Oct 15, 2024
7df068d
Add more tests to xfail set
vivekkhandelwal1 Oct 15, 2024
64f965a
Update abstract_interp_lib_gen.py
vivekkhandelwal1 Oct 15, 2024
3319f4d
Fix fx_importer tests
vivekkhandelwal1 Oct 16, 2024
0e6d87e
Address PR comments
vivekkhandelwal1 Oct 17, 2024
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
24 changes: 24 additions & 0 deletions include/torch-mlir/Dialect/Torch/IR/GeneratedTorchOps.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6317,6 +6317,30 @@ def Torch_AtenDotOp : Torch_Op<"aten.dot", [
let hasCanonicalizer = 1;
}

def Torch_AtenOuterOp : Torch_Op<"aten.outer", [
AllowsTypeRefinement,
HasValueSemantics,
ReadOnly
]> {
let summary = "Generated op for `aten::outer : (Tensor, Tensor) -> (Tensor)`";
let arguments = (ins
AnyTorchTensorType:$self,
AnyTorchTensorType:$vec2
);
let results = (outs
AnyTorchOptionalTensorType:$result
);
let hasCustomAssemblyFormat = 1;
let extraClassDefinition = [{
ParseResult AtenOuterOp::parse(OpAsmParser &parser, OperationState &result) {
return parseDefaultTorchOp(parser, result, 2, 1);
}
void AtenOuterOp::print(OpAsmPrinter &printer) {
printDefaultTorchOp(printer, *this, 2, 1);
}
}];
}

def Torch_AtenCosineSimilarityOp : Torch_Op<"aten.cosine_similarity", [
AllowsTypeRefinement,
HasValueSemantics,
Expand Down
66 changes: 24 additions & 42 deletions lib/Dialect/Torch/Transforms/AbstractInterpLibrary.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7601,6 +7601,13 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
" } : (!torch.int, !torch.bool) -> ()\n"
" return %0 : !torch.list<int>\n"
" }\n"
" func.func @\"__torch_mlir_shape_fn.aten.outer\"(%arg0: !torch.list<int>, %arg1: !torch.list<int>) -> !torch.list<int> {\n"
" %int0 = torch.constant.int 0\n"
" %0 = torch.aten.__getitem__.t %arg0, %int0 : !torch.list<int>, !torch.int -> !torch.int\n"
" %1 = torch.aten.__getitem__.t %arg1, %int0 : !torch.list<int>, !torch.int -> !torch.int\n"
" %2 = torch.prim.ListConstruct %0, %1 : (!torch.int, !torch.int) -> !torch.list<int>\n"
" return %2 : !torch.list<int>\n"
" }\n"
" func.func @\"__torch_mlir_shape_fn.aten.dot\"(%arg0: !torch.list<int>, %arg1: !torch.list<int>) -> !torch.list<int> {\n"
" %0 = torch.prim.ListConstruct : () -> !torch.list<int>\n"
" return %0 : !torch.list<int>\n"
Expand Down Expand Up @@ -13403,6 +13410,14 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
" %4 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%2, %3) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %4 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.outer\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>) -> !torch.int {\n"
" %0:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %1:2 = torch.prim.TupleUnpack %arg1 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %2 = torch.prim.ListConstruct %0#0, %1#0 : (!torch.int, !torch.int) -> !torch.list<optional<int>>\n"
" %3 = torch.prim.ListConstruct %0#1, %1#1 : (!torch.int, !torch.int) -> !torch.list<int>\n"
" %4 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%2, %3) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %4 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.mm\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>) -> !torch.int {\n"
" %false = torch.constant.bool false\n"
" %int5 = torch.constant.int 5\n"
Expand Down Expand Up @@ -13813,63 +13828,30 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
" return %5 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.lerp.Scalar\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>, %arg2: !torch.number) -> !torch.int {\n"
" %none = torch.constant.none\n"
" %0:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %1:2 = torch.prim.TupleUnpack %arg1 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %2 = torch.prim.ListConstruct %0#0, %1#0, %none : (!torch.int, !torch.int, !torch.none) -> !torch.list<optional<int>>\n"
" %3 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.get_dtype_of_scalar(%arg2) : (!torch.number) -> !torch.int\n"
" %4 = torch.prim.ListConstruct %0#1, %1#1, %3 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>\n"
" %5 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%2, %4) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %5 : !torch.int\n"
" %2 = torch.prim.ListConstruct %0#0, %1#0 : (!torch.int, !torch.int) -> !torch.list<optional<int>>\n"
" %3 = torch.prim.ListConstruct %0#1, %1#1 : (!torch.int, !torch.int) -> !torch.list<int>\n"
" %4 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%2, %3) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %4 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.addcmul\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>, %arg2: !torch.tuple<int, int>, %arg3: !torch.number) -> !torch.int {\n"
" %none = torch.constant.none\n"
" %str = torch.constant.str \"AssertionError: \"\n"
" %int11 = torch.constant.int 11\n"
" %0:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %1:2 = torch.prim.TupleUnpack %arg1 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %2:2 = torch.prim.TupleUnpack %arg2 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %3 = torch.aten.ne.int %0#1, %int11 : !torch.int, !torch.int -> !torch.bool\n"
" torch.prim.If %3 -> () {\n"
" torch.prim.If.yield\n"
" } else {\n"
" torch.prim.RaiseException %str, %none : !torch.str, !torch.none\n"
" torch.prim.If.yield\n"
" }\n"
" %4 = torch.aten.ne.int %1#1, %int11 : !torch.int, !torch.int -> !torch.bool\n"
" torch.prim.If %4 -> () {\n"
" torch.prim.If.yield\n"
" } else {\n"
" torch.prim.RaiseException %str, %none : !torch.str, !torch.none\n"
" torch.prim.If.yield\n"
" }\n"
" %5 = torch.aten.ne.int %2#1, %int11 : !torch.int, !torch.int -> !torch.bool\n"
" torch.prim.If %5 -> () {\n"
" torch.prim.If.yield\n"
" } else {\n"
" torch.prim.RaiseException %str, %none : !torch.str, !torch.none\n"
" torch.prim.If.yield\n"
" }\n"
" %6 = torch.prim.ListConstruct %0#0, %1#0, %2#0 : (!torch.int, !torch.int, !torch.int) -> !torch.list<optional<int>>\n"
" %7 = torch.prim.ListConstruct %0#1, %1#1, %2#1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>\n"
" %8 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%6, %7) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %8 : !torch.int\n"
" %3 = torch.prim.ListConstruct %0#0, %1#0, %2#0 : (!torch.int, !torch.int, !torch.int) -> !torch.list<optional<int>>\n"
" %4 = torch.prim.ListConstruct %0#1, %1#1, %2#1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>\n"
" %5 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%3, %4) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" return %5 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.addcdiv\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>, %arg2: !torch.tuple<int, int>, %arg3: !torch.number) -> !torch.int {\n"
" %int6 = torch.constant.int 6\n"
" %0:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %1:2 = torch.prim.TupleUnpack %arg1 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %2:2 = torch.prim.TupleUnpack %arg2 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
" %3 = torch.prim.ListConstruct %0#0, %1#0, %2#0 : (!torch.int, !torch.int, !torch.int) -> !torch.list<optional<int>>\n"
" %4 = torch.prim.ListConstruct %0#1, %1#1, %2#1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>\n"
" %5 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.promote_dtypes(%3, %4) : (!torch.list<optional<int>>, !torch.list<int>) -> !torch.int\n"
" %6 = call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.is_integer_dtype(%5) : (!torch.int) -> !torch.bool\n"
" %7 = torch.prim.If %6 -> (!torch.int) {\n"
" torch.prim.If.yield %int6 : !torch.int\n"
" } else {\n"
" torch.prim.If.yield %5 : !torch.int\n"
" }\n"
" return %7 : !torch.int\n"
" return %5 : !torch.int\n"
" }\n"
" func.func @\"__torch_mlir_dtype_fn.aten.add.Scalar\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.number, %arg2: !torch.number) -> !torch.int {\n"
" %none = torch.constant.none\n"
Expand Down
101 changes: 74 additions & 27 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 @@ -442,10 +442,6 @@
"ElementwiseDequantizePerTensorModule_basic",
"ElementwiseQuantizePerTensorModule_basic",
"ElementwiseQuantizePerTensorUIntModule_basic",
"ElementwiseRreluEvalModule_basic",
"ElementwiseRreluEvalStaticModule_basic",
"ElementwiseRreluTrainModule_basic",
"ElementwiseRreluTrainStaticModule_basic",
"ElementwiseToDtypeI64ToUI8Module_basic",
"EqIntModule_basic",
"FloatImplicitModule_basic",
Expand Down Expand Up @@ -487,9 +483,6 @@
"ReduceMinAlongDimUnsignedInt_basic",
"RsubInt0d_NumToTensor_Module_basic",
"ScalarImplicitFloatModule_basic",
"SignAndLogarithmOfDeterminantModule_F32",
"SignAndLogarithmOfDeterminantBatchedModule_F32",
"SignAndLogarithmOfDeterminantDynamicModule_F32",
"SortIntListReverse_basic",
"SortIntList_basic",
"SplitDimDynamicModule_basic",
Expand Down Expand Up @@ -519,13 +512,42 @@
"SplitTensorNegativeDimModule_basic",
"SplitWithSizesListUnpackModule_basic",
"SplitWithSizes_Module_basic",
"AdaptiveAvgPool1dGeneralDynamic_basic",
"AdaptiveAvgPool1dStaticEvenMultiple_basic",
"AdaptiveAvgPool1dStaticLargerOutput_basic",
"AdaptiveAvgPool2dDynamicNoBatch_basic",
"AdaptiveAvgPool2dDynamic_basic",
"AdaptiveMaxPool1dDynamicNoBatch_basic",
"AdaptiveMaxPool1dDynamic_basic",
"AdaptiveMaxPool1dStatic_basic",
"CrossEntropyLossModule_basic",
"CrossEntropyLossNoReductionModule_basic",
"ElementwiseExpm1IntModule_basic",
"ElementwiseExpm1Module_basic",
"IndexPutImpl1DFloatAccumulateModule_basic",
"IndexPutImpl1DFloatNonAccumulateModule_basic",
"IndexPutImpl1DIntAccumulateModule_basic",
"IndexPutImpl1DIntNonAccumulateModule_basic",
"IndexPutImpl2DFloatNonAccumulateModule_basic",
"IndexPutImpl2DImplicitModule_basic",
"IndexPutImpl2DIndexModule_basic",
"IndexPutImpl2DNoneIndexStaticModule_basic",
"IndexPutImpl3DFloatNonAccumulateModule_basic",
"IndexPutImplIndexWithNoneModule_basic",
"InterpolateDynamicModule_sizes_nearest",
"IouOfModule_basic",
"MeshgridIndexingIJ_basic",
"MeshgridIndexingXY_basic",
"Meshgrid_basic",
"OneHotModule_basic",
}

FX_IMPORTER_CRASHING_SET = LINALG_CRASHING_SET | {
"HBC_basic",
# Runtime op verification: out-of-bounds access
"_SoftmaxModule_basic",
"UpSampleNearest2dDynamicFactor_basic",
"AdaptiveAvgPool1dGeneralDynamicNoBatches_basic",
}

FX_IMPORTER_STABLEHLO_XFAIL_SET = {
Expand Down Expand Up @@ -554,10 +576,6 @@
"ElementwiseRemainderTensorModule_Int_Float_NegativeDivisor_basic",
"ElementwiseRemainderTensorModule_Int_NegativeDividend_basic",
"ElementwiseRemainderTensorModule_Int_NegativeDivisor_basic",
"ElementwiseRreluEvalModule_basic",
"ElementwiseRreluEvalStaticModule_basic",
"ElementwiseRreluTrainModule_basic",
"ElementwiseRreluTrainStaticModule_basic",
"MaxPool1dCeilModeTrueModule_basic",
"MaxPool1dStaticCeilModeTrueModule_basic",
"MaxUnpool3dModulePad0_basic",
Expand Down Expand Up @@ -591,7 +609,6 @@
"AdaptiveAvgPool3dDynamic_basic",
"AdaptiveMaxPool1dDynamicNoBatch_basic",
"AdaptiveMaxPool1dDynamic_basic",
"AdaptiveMaxPool1dDimOneStatic_basic",
"AdaptiveMaxPool1dStatic_basic",
"AdaptiveMaxPool2dDynamicNoBatch_basic",
"AdaptiveMaxPool2dDynamicWithIndices_basic",
Expand Down Expand Up @@ -758,12 +775,7 @@
"MaxPool2dWithIndicesBackwardStatic3DModule_basic",
"MaxPool2dWithIndicesBackwardStatic4DModule_basic",
"MaxPool3dCeilModeTrueModule_basic",
"MaxPool3dEmptyStrideStaticModule_basic",
"MaxPool3dLargeDatadModule_basic",
"MaxPool3dModuleRandomSimple_basic",
"MaxPool3dModule_basic",
"MaxPool3dStaticCeilModeTrueModule_basic",
"MaxPool3dStaticModule_basic",
"MaxPool3dWithIndicesAllNegativeValuesModule_basic",
"MaxPool3dWithIndicesAllOnesModule_basic",
"MaxPool3dWithIndicesCeilModeTrueModule_basic",
Expand Down Expand Up @@ -921,6 +933,51 @@
"Unfold_Module_Rank_Zero_basic",
"Unfold_Module_Rank_Zero_Size_Zero_basic",
"Unfold_Module_Dynamic_basic",
"AdaptiveAvgPool1dGeneralDynamic_basic",
"AdaptiveAvgPool1dGeneralDynamicNoBatches_basic",
"AdaptiveAvgPool1dStaticEvenMultiple_basic",
"AdaptiveAvgPool1dStaticLargerOutput_basic",
"AdaptiveAvgPool2dDynamicNoBatch_basic",
"AdaptiveAvgPool2dDynamic_basic",
"AddIntModule_basic",
"AtenIntTensorByteDtypeModule_basic",
"AtenIntTensorCharDtypeModule_basic",
"AtenItemIntOpModule_basic",
"CrossEntropyLossModule_basic",
"CrossEntropyLossNoReductionModule_basic",
"EinsumStaticContractRhsModule_basic",
"EinsumStaticFourDimensionModule_basic",
"EinsumStaticModule_basic",
"EinsumStaticWithEllipsisSlicingAndBroadcastModule_basic",
"EinsumStaticWithEllipsisSlicingModule_basic",
"ElementwiseExpm1IntModule_basic",
"ElementwiseExpm1Module_basic",
"InterpolateDynamicModule_sizes_nearest",
"IouOfModule_basic",
"IscloseStaticModuleTrue_basic",
"IscloseStaticModule_basic",
"MeshgridIndexingIJ_basic",
"MeshgridIndexingXY_basic",
"Meshgrid_basic",
"MulIntModule_basic",
"OneHotModule_basic",
"ReduceFrobeniusNormComplexModule_basic",
"ScalarImplicitIntModule_basic",
"ScaledDotProductAttentionBoolMaskModule_basic",
"ScaledDotProductAttentionDifferentCausalModule_basic",
"ScaledDotProductAttentionDifferentDynamicCausalModule_basic",
"ScaledDotProductAttentionDifferentModule_basic",
"ScaledDotProductAttentionMaskModule_basic",
"ScaledDotProductAttentionSameCausalModule_basic",
"ScaledDotProductAttentionSameDynamicModule_basic",
"ScaledDotProductAttentionSameModule_basic",
"SubIntModule_basic",
"TensorToIntZeroRank_basic",
"UpSampleNearest2dDynamicFactor_basic",
"UpSampleNearest2dDynamicSize_basic",
"UpSampleNearest2dStaticFactor_basic",
"UpSampleNearest2dStaticSize_basic",
"UpSampleNearest2d_basic",
}

FX_IMPORTER_STABLEHLO_CRASHING_SET = {
Expand Down Expand Up @@ -3297,7 +3354,6 @@
"SplitWithSizesListUnpackModule_basic",
"SplitWithSizes_Module_basic",
"ElementwiseCreateComplexModule_basic",
"AdaptiveMaxPool1dDimOneStatic_basic",
"AtenPolarDoubleModule_basic",
"AtenPolarFloatModule_basic",
"HstackBasicComplexModule_basic",
Expand All @@ -3318,10 +3374,6 @@
"Conv_Transpose3dStaticModule_basic",
"ElementwiseFloatTensorGtIntTensorModule_basic",
"ElementwiseIntTensorLtFloatTensorModule_basic",
"ElementwiseRreluEvalModule_basic",
"ElementwiseRreluEvalStaticModule_basic",
"ElementwiseRreluTrainModule_basic",
"ElementwiseRreluTrainStaticModule_basic",
"IndexPutWithNoneAndBroadcastModule_basic",
"MaskedScatterStaticBasic_basic",
"MaxUnpool3dModulePad0_basic",
Expand Down Expand Up @@ -3628,12 +3680,7 @@
"MaxPool2dWithIndicesNonDefaultStrideModule_basic",
"MaxPool2dWithIndicesStaticModule_basic",
"MaxPool3dCeilModeTrueModule_basic",
"MaxPool3dEmptyStrideStaticModule_basic",
"MaxPool3dLargeDatadModule_basic",
"MaxPool3dModuleRandomSimple_basic",
"MaxPool3dModule_basic",
"MaxPool3dStaticCeilModeTrueModule_basic",
"MaxPool3dStaticModule_basic",
"MaxPool3dWithIndicesAllNegativeValuesModule_basic",
"MaxPool3dWithIndicesAllOnesModule_basic",
"MaxPool3dWithIndicesCeilModeTrueModule_basic",
Expand Down
Loading
Loading