Skip to content

Add CUDA implementation of GatherBlockQuantized operator #25575

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

Merged
merged 24 commits into from
Aug 1, 2025
+417 −4
Merged

Add CUDA implementation of GatherBlockQuantized operator #25575

Changes from 1 commit
Commits
Show all changes
24 commits
Select commit Hold shift + click to select a range
7a63bfa
Add CUDA implementation of GatherBlockQuantized operator
Jul 30, 2025
4256d1b
Clang format
Jul 30, 2025
ec4c2d3
Run with lint runner
Jul 30, 2025
5d159c2
Add indices shape check
Jul 30, 2025
5bf1c91
Fix webgpu tests
Jul 30, 2025
d992d72
Disable webgpu tests which are failing
Jul 31, 2025
8f16bac
remove unused formal parameter
Jul 31, 2025
fc31dc4
void unused parameter
Jul 31, 2025
cbad96f
void unused parameter
Jul 31, 2025
20ed595
Fix nodiscard status
Jul 31, 2025
9a39521
Fix integer precision
Jul 31, 2025
92da3cd
cast raw data instead of using template for data fetching
Jul 31, 2025
941dd73
Fix CUDA_MS_OP_THREE_TYPED_CLASS_NAME
Jul 31, 2025
eb4522d
Fix mismatched marcro name
Jul 31, 2025
ef55de8
Fix BuildKernelCreateInfo
Jul 31, 2025
7167c2d
Use raw data
Jul 31, 2025
75e497b
Fix typo
Jul 31, 2025
b023cd8
Fix axis check
Jul 31, 2025
15c3219
Fix kernel
Aug 1, 2025
2f73aac
Support both signed and unsigned data type
Aug 1, 2025
a702037
Fix templates
Aug 1, 2025
eab73b5
Fix sign type
Aug 1, 2025
0f67c22
Disable more tests
Aug 1, 2025
0c7938e
Update autogenerated doc
Aug 1, 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
Prev Previous commit
Next Next commit
Disable more tests
  • Loading branch information
Xiaoyan Hu committed Aug 1, 2025
commit 0f67c22ef3f5ecaf010770b31e20131f9c8eab5d
10 changes: 10 additions & 0 deletions onnxruntime/test/contrib_ops/gather_block_quantized_op_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -280,6 +280,7 @@ void Test_Fail_WithZeroPoints(int64_t gather_axis,
gather_axis, quantize_axis, block_size, bits, output, output_shape, false);
}

#ifndef USE_CUDA
TEST(GatherBlockQuantizedOpTest, UnsupportedTypes) {
Test_Fail_WithZeroPoints<int8_t, float, int32_t>(0, 2, 16);
Test_Fail_WithZeroPoints<int16_t, float, int32_t>(0, 2, 16);
Expand All @@ -294,6 +295,7 @@ TEST(GatherBlockQuantizedOpTest, UnsupportedTypes) {
Test_Fail_WithZeroPoints<Int4x2, BFloat16, int32_t>(0, 2, 16);
Test_Fail_WithZeroPoints<uint8_t, float, int16_t>(0, 2, 16);
}
#endif

template <typename T1, typename T2, typename Tind>
void Test_Fail_WithoutZeroPoints(int64_t gather_axis,
Expand Down Expand Up @@ -322,6 +324,7 @@ void Test_Fail_WithoutZeroPoints(int64_t gather_axis,
gather_axis, quantize_axis, block_size, bits, output, output_shape, false);
}

#ifndef USE_CUDA
TEST(GatherBlockQuantizedOpTest, UnsupportedUInt8DataType) {
// Gather on axis other than 0 is not supported with uint8_t
Test_Fail_WithoutZeroPoints<uint8_t, float, int32_t>(1, 2, 16);
Expand Down Expand Up @@ -354,6 +357,7 @@ TEST(GatherBlockQuantizedOpTest, NotSupportedBits) {
Test_Fail_WithZeroPoints<UInt4x2, float, int32_t>(0, 2, 16, 6);
Test_Fail_WithZeroPoints<UInt4x2, float, int32_t>(0, 2, 16, 7);
}
#endif

template <typename T1, typename T2, typename Tind>
void Test_ShapeMismatch_WithZeroPoints() {
Expand Down Expand Up @@ -382,11 +386,13 @@ void Test_ShapeMismatch_WithZeroPoints() {
gather_axis, quantize_axis, block_size, bits, output, output_shape, false);
}

#ifndef USE_CUDA
TEST(GatherBlockQuantizedOpTest, ShapeMismatch) {
Test_ShapeMismatch_WithZeroPoints<UInt4x2, float, int32_t>();
Test_ShapeMismatch_WithZeroPoints<Int4x2, float, int32_t>();
Test_ShapeMismatch_WithZeroPoints<uint8_t, float, int32_t>();
}
#endif

template <typename T1, typename T2, typename Tind>
void Test_InvalidIndices_WithZeroPoints() {
Expand Down Expand Up @@ -415,11 +421,13 @@ void Test_InvalidIndices_WithZeroPoints() {
gather_axis, quantize_axis, block_size, bits, output, output_shape, false, true);
}

#ifndef USE_CUDA
TEST(GatherBlockQuantizedOpTest, InvalidIndices) {
Test_InvalidIndices_WithZeroPoints<UInt4x2, float, int32_t>();
Test_InvalidIndices_WithZeroPoints<Int4x2, float, int32_t>();
Test_InvalidIndices_WithZeroPoints<uint8_t, float, int32_t>();
}
#endif

template <typename T1, typename T2, typename Tind>
void Test_GatherAxis0_WithZeroPoints(int bits = 4) {
Expand Down Expand Up @@ -452,6 +460,7 @@ void Test_GatherAxis0_WithZeroPoints(int bits = 4) {
-3, -1, block_size, bits, output, output_shape, true);
}

#ifndef USE_CUDA
TEST(GatherBlockQuantizedOpTest, GatherAxis0WithZeroPoints) {
Test_GatherAxis0_WithZeroPoints<UInt4x2, float, int32_t>();
Test_GatherAxis0_WithZeroPoints<Int4x2, float, int32_t>();
Expand All @@ -462,6 +471,7 @@ TEST(GatherBlockQuantizedOpTest, GatherAxis0WithZeroPoints) {
Test_GatherAxis0_WithZeroPoints<UInt4x2, MLFloat16, int64_t>();
Test_GatherAxis0_WithZeroPoints<Int4x2, MLFloat16, int64_t>();
}
#endif

template <typename T1, typename T2, typename Tind>
void Test_GatherAxis0_WithZeroPoints_Uint8(int bits = 4) {
Expand Down