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

Add support for unsigned types with mlir #3582

Open
wants to merge 9 commits into
base: develop
Choose a base branch
from
+224 −38
Open

Add support for unsigned types with mlir #3582

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
e01065e
Add signed/unsigned type
pfultz2 Oct 31, 2024
ffc242f
Update int4 tests
pfultz2 Oct 31, 2024
9dd0e23
Add unsigned types
pfultz2 Oct 31, 2024
7b458c0
Add unsigned tests
pfultz2 Nov 10, 2024
9c25af0
Add unsigned dq test
pfultz2 Nov 10, 2024
4200ad9
Format
pfultz2 Nov 10, 2024
65c34cb
Merge branch 'develop' into mlir-unsigned
pfultz2 Nov 10, 2024
21225be
Merge branch 'develop' into mlir-unsigned
causten Nov 13, 2024
75e4348
Use attribute variable instead of hard-coding them
dhernandez0 Nov 15, 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
2 changes: 2 additions & 0 deletions src/targets/gpu/fuse_mlir.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -366,7 +366,9 @@ bool is_pointwise_op_supported_by_mlir(const instruction& i)
type_t::fp8e4m3fn_type,
type_t::fp8e5m2_type,
type_t::int8_type,
type_t::uint8_type,
type_t::int32_type,
type_t::uint32_type,
type_t::bool_type};
// Preliminary type check.
if(not contains(allowed_types, result_type))
Expand Down
21 changes: 5 additions & 16 deletions src/targets/gpu/mlir.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -322,14 +322,14 @@ struct mlir_program
result = mlirF64TypeGet(ctx.get());
else if(as.is_integral())
{
// Note: rocMLIR use signless integer type for tensors types. This
// will translate to signed implementation for current supported
// operations.
if(as.is_unsigned())
{
MIGRAPHX_THROW("Unsupported type: " + std::to_string(as.type_enum()));
result = mlirIntegerTypeUnsignedGet(ctx.get(), as.size() * 8);
}
else
{
result = mlirIntegerTypeSignedGet(ctx.get(), as.size() * 8);
}
result = mlirIntegerTypeGet(ctx.get(), as.size() * 8); // number of bits
}
else
MIGRAPHX_THROW("Unsupported type: " + std::to_string(as.type_enum()));
Expand Down Expand Up @@ -718,24 +718,13 @@ struct mlir_program
literal r = ins->get_literal();
auto sh = ins->get_shape();

// mlir works only with signed types. change uint4 to (int4 + unsigned-flag)
if(shape::is_unsigned(sh.type()) and ins->outputs()[0]->name() == "unpack_int4")
sh = ins->get_shape().with_type(shape::int8_type);

MlirType shaped_type = make_mlir_shaped(sh);
MlirType tensor_type = rocmlirMIXRShapedTypeAsTensor(shaped_type);
MlirAttribute mlir_value_attr =
mlirDenseElementsAttrRawBufferGet(tensor_type, r.get_shape().bytes(), r.data());
ops.add_attributes({{"value", mlir_value_attr}});
}

if(ins->name() == "unpack_int4")
{
auto sh = get_shape(ins);
ops.add_attributes(
{{"isUnsigned", shape::is_unsigned(sh.type())}}); // flag for uint4
}

if(ins->name() == "convolution" or ins->name() == "dot")
{
pp =
Expand Down
83 changes: 83 additions & 0 deletions test/gpu/fuse_mlir.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -528,6 +528,89 @@ TEST_CASE(dequantizelinear_dot)
EXPECT(p1.sort() == p2.sort());
}

TEST_CASE(unsigned_dequantizelinear_dot)
{
migraphx::program p1;
{
auto* mm = p1.get_main_module();

auto x = mm->add_parameter("x", migraphx::shape{migraphx::shape::float_type, {2, 3, 5}});
auto y = mm->add_parameter("y", migraphx::shape{migraphx::shape::uint8_type, {2, 5, 2}});
auto scalelit =
mm->add_literal(migraphx::generate_literal({migraphx::shape::float_type, {2, 2, 2}}));
auto zplit =
mm->add_literal(migraphx::generate_literal({migraphx::shape::uint8_type, {2, 2, 2}}));

auto unsqueeze1 =
mm->add_instruction(migraphx::make_op("unsqueeze", {{"axes", {2}}}), scalelit);
auto broadcast1 = mm->add_instruction(
migraphx::make_op("multibroadcast", {{"out_lens", {2, 2, 3, 2}}}), unsqueeze1);
auto reshape1 =
mm->add_instruction(migraphx::make_op("reshape", {{"dims", {2, 6, 2}}}), broadcast1);
auto scale = mm->add_instruction(
migraphx::make_op("slice", {{"axes", {1}}, {"starts", {0}}, {"ends", {5}}}), reshape1);

auto unsqueeze2 =
mm->add_instruction(migraphx::make_op("unsqueeze", {{"axes", {2}}}), zplit);
auto broadcast2 = mm->add_instruction(
migraphx::make_op("multibroadcast", {{"out_lens", {2, 2, 3, 2}}}), unsqueeze2);
auto reshape2 =
mm->add_instruction(migraphx::make_op("reshape", {{"dims", {2, 6, 2}}}), broadcast2);
auto zp = mm->add_instruction(
migraphx::make_op("slice", {{"axes", {1}}, {"starts", {0}}, {"ends", {5}}}), reshape2);

auto dq = add_pointwise(
p1, "main:pointwise0", {y, scale, zp}, single_pointwise("dequantizelinear"));
auto dot = mm->add_instruction(migraphx::make_op("dot"), x, dq);
mm->add_return({dot});
}
run_pass(p1);
migraphx::program p2;
{
auto* mm = p2.get_main_module();
auto x = mm->add_parameter("x", migraphx::shape{migraphx::shape::float_type, {2, 3, 5}});
auto y = mm->add_parameter("y", migraphx::shape{migraphx::shape::uint8_type, {2, 5, 2}});
auto scalelit =
mm->add_literal(migraphx::generate_literal({migraphx::shape::float_type, {2, 2, 2}}));
auto zplit =
mm->add_literal(migraphx::generate_literal({migraphx::shape::uint8_type, {2, 2, 2}}));

auto fused = add_mlir(
p2,
"main:pointwise0:mlir_dot0",
{y, scalelit, zplit, x},
{"x0", "x1", "x2", "x3"},
[=](auto* pm, const auto& inputs) {
auto unsqueeze1 =
pm->add_instruction(migraphx::make_op("unsqueeze", {{"axes", {2}}}), inputs[1]);
auto broadcast1 = pm->add_instruction(
migraphx::make_op("multibroadcast", {{"out_lens", {2, 2, 3, 2}}}), unsqueeze1);
auto reshape1 = pm->add_instruction(
migraphx::make_op("reshape", {{"dims", {2, 6, 2}}}), broadcast1);
auto scale = pm->add_instruction(
migraphx::make_op("slice", {{"axes", {1}}, {"starts", {0}}, {"ends", {5}}}),
reshape1);

auto unsqueeze2 =
pm->add_instruction(migraphx::make_op("unsqueeze", {{"axes", {2}}}), inputs[2]);
auto broadcast2 = pm->add_instruction(
migraphx::make_op("multibroadcast", {{"out_lens", {2, 2, 3, 2}}}), unsqueeze2);
auto reshape2 = pm->add_instruction(
migraphx::make_op("reshape", {{"dims", {2, 6, 2}}}), broadcast2);
auto zp = pm->add_instruction(
migraphx::make_op("slice", {{"axes", {1}}, {"starts", {0}}, {"ends", {5}}}),
reshape2);

auto dq = pm->add_instruction(
migraphx::make_op("dequantizelinear"), inputs[0], scale, zp);
auto dot = pm->add_instruction(migraphx::make_op("dot"), inputs[3], dq);
return std::make_tuple(dot->get_operator(), dot);
});
mm->add_return({fused});
}
EXPECT(p1.sort() == p2.sort());
}

TEST_CASE(unpack_int4_dot)
{
migraphx::program p1;
Expand Down
Loading
Loading