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

Jump to bottom

[ONNX] Fix resize ceil numerics and add half_pixel_symmetric support #3443

Merged
merged 3 commits into from
Jun 12, 2024
Merged

[ONNX] Fix resize ceil numerics and add half_pixel_symmetric support #3443

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c9a52cb
Fix resize ceil numerics and add support for half_pixel_symmetric
zjgarvey Jun 10, 2024
c5f21bc
Merge remote-tracking branch 'upstream/main' into resize_ceil_fix
zjgarvey Jun 10, 2024
2ecb77e
add some lit tests
zjgarvey Jun 11, 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
22 changes: 21 additions & 1 deletion lib/Conversion/TorchToLinalg/Uncategorized.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2657,14 +2657,21 @@ static Value NearestInterpolate(OpBuilder &b, Location loc,
nearestFP = b.create<arith::SelectOp>(loc, cmp, floor, ceil);
} else if (nearestMode == "round_prefer_ceil") {
Value cstHalf = b.create<arith::ConstantOp>(loc, b.getF32FloatAttr(0.5));
Value cstOne = b.create<arith::ConstantOp>(loc, b.getF32FloatAttr(1));
Value floor = b.create<math::FloorOp>(loc, proj);
Value ceil = b.create<math::CeilOp>(loc, proj);
Value decimal = b.create<arith::SubFOp>(loc, proj, floor);
Value cmp = b.create<arith::CmpFOp>(loc, arith::CmpFPredicate::UGE,
decimal, cstHalf);
nearestFP = b.create<arith::SelectOp>(loc, cmp, ceil, floor);
Value inputSizeMOne = b.create<arith::SubFOp>(loc, inputSizeFP, cstOne);
// don't extract out of bounds
nearestFP = b.create<arith::MinimumFOp>(loc, nearestFP, inputSizeMOne);
} else if (nearestMode == "ceil") {
Value cstOne = b.create<arith::ConstantOp>(loc, b.getF32FloatAttr(1));
Value inputSizeMOne = b.create<arith::SubFOp>(loc, inputSizeFP, cstOne);
nearestFP = b.create<math::CeilOp>(loc, proj);
nearestFP = b.create<arith::MinimumFOp>(loc, nearestFP, inputSizeMOne);
} else {
llvm_unreachable("Unsupported nearest mode");
}
Expand Down Expand Up @@ -2738,7 +2745,8 @@ static Value BilinearInterpolate(OpBuilder &b,
if (coordStr == "_asymmetric") {
preClip = b.create<arith::DivFOp>(loc, outFP, scale);
}
if (coordStr == "_pytorch_half_pixel" || coordStr == "") {
if (coordStr == "_pytorch_half_pixel" || coordStr == "" ||
coordStr == "_half_pixel_symmetric") {
// half-pixel modes
// y_resized + 0.5
Value outPlusHalf = b.create<arith::AddFOp>(loc, outFP, cstHalf);
Expand All @@ -2747,6 +2755,18 @@ static Value BilinearInterpolate(OpBuilder &b,
// _ - 0.5
preClip = b.create<arith::SubFOp>(loc, outDivScale, cstHalf);
}
// for half_pixel_symmetric, need to compute offset from raw scales
if (coordStr == "_half_pixel_symmetric" && !scaleValues.empty()) {
Value outputSizeFromScale = b.create<arith::MulFOp>(loc, inputFP, scale);
Value adjustment =
b.create<arith::DivFOp>(loc, outputSizeFP, outputSizeFromScale);
Value cstTwo = b.create<arith::ConstantOp>(loc, b.getF32FloatAttr(2.0));
Value center = b.create<arith::DivFOp>(loc, inputFP, cstTwo);
Value oneMAdjustment =
b.create<arith::SubFOp>(loc, cstOneFloat, adjustment);
Value offset = b.create<arith::MulFOp>(loc, center, oneMAdjustment);
preClip = b.create<arith::AddFOp>(loc, offset, preClip);
}
// for pytorch half pixel , special case for length_resized == 1:
if (coordStr == "_pytorch_half_pixel") {
Value cmp = b.create<arith::CmpFOp>(loc, arith::CmpFPredicate::UEQ,
Expand Down
84 changes: 83 additions & 1 deletion test/Conversion/TorchToLinalg/resize.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,7 +156,89 @@ func.func @test_resize_nearest_3d(%arg0: !torch.vtensor<[?,?,?,?,?],f32>, %arg1:
return %7 : !torch.vtensor<[?,?,?,?,?],f32>
}

// CHECK-LABEL: func.func @test_resize_nearest_half_pixel
// -----

// CHECK-LABEL: func.func @test_resize_nearest_ceil
func.func @test_resize_nearest_ceil(%arg0: !torch.vtensor<[?,?,?],f32>, %arg1: !torch.vtensor<[3],si64>) -> !torch.vtensor<[?,?,?],f32> {
// CHECK: %[[GENERIC:.*]] = linalg.generic
// CHECK: %[[x11:.*]] = linalg.index 0 : index
// CHECK: %[[x12:.*]] = linalg.index 1 : index
// CHECK: %[[x13:.*]] = linalg.index 2 : index
// CHECK: %[[x15:.*]] = arith.sitofp %[[c2_i64:.*]] : i64 to f32
// CHECK: %[[x19:.*]] = arith.sitofp %[[x6:.*]] : i64 to f32
// CHECK: %[[x21:.*]] = arith.divf %[[x19]], %[[x15]] : f32
// CHECK: %[[x23:.*]] = arith.index_cast %[[x13]] : index to i64
// CHECK: %[[x24:.*]] = arith.sitofp %[[x23]] : i64 to f32
// CHECK: %[[cst:.*]] = arith.constant 5.000000e-01 : f32
// CHECK: %[[add:.*]] = arith.addf %[[x24]], %[[cst]] : f32
// CHECK: %[[x25:.*]] = arith.divf %[[add]], %[[x21]] : f32
// CHECK: %[[sub:.*]] = arith.subf %[[x25]], %[[cst]] : f32
// CHECK: %[[cst3:.*]] = arith.constant 1.000000e+00 : f32
// CHECK: %[[nM1:.*]] = arith.subf %[[inputsizefp:.*]], %[[cst3]]
// CHECK: %[[ceil:.*]] = math.ceil %[[sub]] : f32
// CHECK: %[[minindex:.*]] = arith.minimumf %[[ceil]], %[[nM1]]
// CHECK: %[[x31:.*]] = arith.fptosi %[[minindex]] : f32 to i64
// CHECK: %[[x32:.*]] = arith.index_cast %[[x31]] : i64 to index
// CHECK: %[[extracted:.*]] = tensor.extract %[[x0:.*]][%[[x11]], %[[x12]], %[[x32]]] : tensor<?x?x?xf32>
// CHECK: linalg.yield %[[extracted]] : f32
%none = torch.constant.none
%none_0 = torch.constant.none
%int0 = torch.constant.int 0
%false = torch.constant.bool false
%true = torch.constant.bool true
%str = torch.constant.str "nearest_half_pixel,ceil"
%int2 = torch.constant.int 2
%0 = torch.aten.select.int %arg1, %int0, %int2 : !torch.vtensor<[3],si64>, !torch.int, !torch.int -> !torch.vtensor<[1],si64>
%1 = torch.aten.item %0 : !torch.vtensor<[1],si64> -> !torch.int
%4 = torch.prim.ListConstruct %1 : (!torch.int) -> !torch.list<int>
%5 = torch.aten.__interpolate.size_list_scale_list %arg0, %4, %none_0, %str, %false, %none_0, %false : !torch.vtensor<[?,?,?],f32>, !torch.list<int>, !torch.none, !torch.str, !torch.bool, !torch.none, !torch.bool -> !torch.vtensor<[?,?,?],f32>
return %5 : !torch.vtensor<[?,?,?],f32>
}

// -----

// CHECK-LABEL: func.func @test_resize_scales_linear_half_pixel_symmetric
func.func @test_resize_scales_linear_half_pixel_symmetric(%arg0: !torch.vtensor<[1,1,2,4],f32>, %arg1: !torch.vtensor<[4]
,f64>) -> !torch.vtensor<[?,?,?,?],f32> attributes {torch.onnx_meta.ir_version = 7 : si64, torch.onnx_meta.opset_version = 19 : si64, torch.onnx_meta.producer_name = "backend-test", torch.onnx_meta.producer_version = ""} {
// CHECK: %[[generic:.*]] = linalg.generic
// CHECK: %[[cst7:.*]] = arith.constant 2.0
// CHECK: %[[halfsize:.*]] = arith.divf %[[sizefp:.*]], %[[cst7]]
// CHECK: %[[modifier:.*]] = arith.subf %[[cstOne:.*]], %[[adjustment:.*]]
// CHECK: %[[offset:.*]] = arith.mulf %[[halfsize]], %[[modifier]]
// CHECK: %[[preClip:.*]] = arith.addf %[[offset]], %[[halfpixelbase:.*]]
// CHECK: %[[extracted:.*]] = tensor.extract %[[x0:.*]][%[[x1:.*]], %[[x2:.*]], %[[x3:.*]], %[[x4:.*]]] : tensor<1x1x2x4xf32>
// CHECK: %[[extracted_7:.*]] = tensor.extract %[[x0]][%[[x1]], %[[x2]]
// CHECK: %[[extracted_8:.*]] = tensor.extract %[[x0]][%[[x1]], %[[x2]]
// CHECK: %[[extracted_9:.*]] = tensor.extract %[[x0]][%[[x1]], %[[x2]]
// CHECK: %[[dx0p00:.*]] = arith.mulf %[[dx0:.*]], %[[extracted]]
// CHECK: %[[dx1p01:.*]] = arith.mulf %[[dx1:.*]], %[[extracted_7]]
// CHECK: %[[sum:.*]] = arith.addf %[[dx0p00]], %[[dx1p01]]
// CHECK: %[[left:.*]] = arith.mulf %[[dy0:.*]], %[[sum]]
// CHECK: %[[dx0p10:.*]] = arith.mulf %[[dx0]], %[[extracted_8]]
// CHECK: %[[dx1p11:.*]] = arith.mulf %[[dx1]], %[[extracted_9]]
// CHECK: %[[sum2:.*]] = arith.addf %[[dx0p10]], %[[dx1p11]]
// CHECK: %[[right:.*]] = arith.mulf %[[dy1:.*]], %[[sum2]]
// CHECK: %[[retval:.*]] = arith.addf %[[left]], %[[right]]
%none = torch.constant.none
%none_0 = torch.constant.none
%int0 = torch.constant.int 0
%false = torch.constant.bool false
%true = torch.constant.bool true
%str = torch.constant.str "bilinear_half_pixel_symmetric"
%int2 = torch.constant.int 2
%0 = torch.aten.select.int %arg1, %int0, %int2 : !torch.vtensor<[4],f64>, !torch.int, !torch.int -> !torch.vtensor<[1],f64>
%1 = torch.aten.item %0 : !torch.vtensor<[1],f64> -> !torch.float
%int3 = torch.constant.int 3
%2 = torch.aten.select.int %arg1, %int0, %int3 : !torch.vtensor<[4],f64>, !torch.int, !torch.int -> !torch.vtensor<[1],f64>
%3 = torch.aten.item %2 : !torch.vtensor<[1],f64> -> !torch.float
%4 = torch.prim.ListConstruct %1, %3 : (!torch.float, !torch.float) -> !torch.list<float>
%5 = torch.aten.__interpolate.size_list_scale_list %arg0, %none_0, %4, %str, %false, %none_0, %false : !torch.vtensor<[1,1,2,4],f32>, !torch.none, !torch.list<float>, !torch.str, !torch.bool, !torch.none, !torch.bool -> !torch.vtensor<[?,?,?,?],f32>
return %5 : !torch.vtensor<[?,?,?,?],f32>
}

// -----

// CHECK-LABEL: func.func @test_resize_nearest_half_pixel_round_prefer_floor
func.func @test_resize_nearest_half_pixel_round_prefer_floor(%arg0: !torch.vtensor<[?,?,?],f32>, %arg1: !torch.vtensor<[3],si64>) -> !torch.vtensor<[?,?,?],f32> {
// CHECK: %[[GENERIC:.*]] = linalg.generic
// CHECK: %[[x11:.*]] = linalg.index 0 : index
Expand Down
Loading