Skip to content

【bug fix】修复w4a8编译慢 #3510

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 3 commits into from
Aug 21, 2025
Merged

【bug fix】修复w4a8编译慢 #3510

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
8fed166
修复w4a8编译
yangjianfengo1 Aug 21, 2025
60f50b8
code style
yangjianfengo1 Aug 21, 2025
3670bcd
修复tma copy
yangjianfengo1 Aug 21, 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
46 changes: 3 additions & 43 deletions custom_ops/gpu_ops/w4afp8_gemm/w4afp8_gemm.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,16 +77,7 @@ void DisPatchW4AFp8Gemm(
max_tokens,
stream)
} else {
GEMM_SWITCH_FP16(
M, K, batch_size, token_padding_size, kBlockN, TailN,
weight,
input,
out,
weight_scale,
input_row_sum,
tokens,
max_tokens,
stream)
PD_THROW("Only supported dtype in ['BFLOAT16'].");
}
}

Expand Down Expand Up @@ -128,22 +119,7 @@ std::vector<paddle::Tensor> W4AFp8Gemm(
input.stream());
return {out};
} else {
paddle::Tensor out = paddle::empty({all_tokens, M}, paddle::DataType::FLOAT16, input.place());
phi::dtype::float16 *out_data = out.data<phi::dtype::float16>();
DisPatchW4AFp8Gemm(
reinterpret_cast<const cutlass::float_e4m3_t*>(input.data<phi::dtype::float8_e4m3fn>()),
reinterpret_cast<const cutlass::float_e4m3_t*>(weight.data<uint8_t>()),
tokens.data<int>(),
input_row_sum.data<float>(),
weight_scale.data<float>(),
reinterpret_cast<cutlass::half_t*>(out_data),
token_padding_size,
max_tokens,
batch_size,
M,
K,
input.stream());
return {out};
PD_THROW("Only supported dtype in ['BFLOAT16'].");
}
} else {
if (is_bflot16) {
Expand All @@ -164,23 +140,7 @@ std::vector<paddle::Tensor> W4AFp8Gemm(
input.stream());
return {out};
} else {
paddle::Tensor out = paddle::empty({batch_size, token_padding_size, M}, paddle::DataType::FLOAT16, input.place());
phi::dtype::float16 * out_data = out.data<phi::dtype::float16>();

DisPatchW4AFp8Gemm(
reinterpret_cast<const cutlass::float_e4m3_t*>(input.data<phi::dtype::float8_e4m3fn>()),
reinterpret_cast<const cutlass::float_e4m3_t*>(weight.data<uint8_t>()),
tokens.data<int>(),
input_row_sum.data<float>(),
weight_scale.data<float>(),
reinterpret_cast<cutlass::half_t*>(out_data),
token_padding_size,
max_tokens,
batch_size,
M,
K,
input.stream());
return {out};
PD_THROW("Only supported dtype in ['BFLOAT16'].");
}
}
}
Expand Down
11 changes: 3 additions & 8 deletions custom_ops/utils/auto_gen_w4afp8_gemm_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,14 +83,9 @@
}}
"""

gemm_case = [
[8192, 3584, 8, 0], # eb45T ffn1
[8192, 3584, 8, 2048], # eb45T ffn1
[7168, 8192, 8, 0], # eb45T ffn2
[7168, 8192, 8, 2048], # eb45T ffn2
]

dtype = ["BF16", "FP16"]
gemm_case = [[256, 256, 1, 0]]

dtype = ["BF16"]


def get_cutlass_type(type):
Expand Down
8 changes: 4 additions & 4 deletions tests/operators/test_w4afp8_gemm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,10 +44,10 @@ def peruate_scale(weight_scale):


paddle.seed(0)
tokens_per_group = 32
N = 8192
K = 3584
BATCH = 8
tokens_per_group = 256
N = 256
K = 256
BATCH = 1
TokenPadding = 0

tokens = [tokens_per_group] * BATCH
Expand Down
Loading