Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
vllm_cscc
Commits
82e0d601
Unverified
Commit
82e0d601
authored
Feb 22, 2025
by
Roger Wang
Committed by
GitHub
Feb 22, 2025
Browse files
[CI/Build] Fix pre-commit errors from #13571 (#13709)
Signed-off-by:
Roger Wang
<
ywang@roblox.com
>
parent
78ac0f59
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
7 additions
and
5 deletions
+7
-5
csrc/quantization/fp4/nvfp4_scaled_mm_entry.cu
csrc/quantization/fp4/nvfp4_scaled_mm_entry.cu
+4
-3
csrc/quantization/fp4/nvfp4_scaled_mm_kernels.cu
csrc/quantization/fp4/nvfp4_scaled_mm_kernels.cu
+3
-2
No files found.
csrc/quantization/fp4/nvfp4_scaled_mm_entry.cu
View file @
82e0d601
...
@@ -31,7 +31,8 @@ void cutlass_scaled_fp4_mm(torch::Tensor& D, torch::Tensor const& A,
...
@@ -31,7 +31,8 @@ void cutlass_scaled_fp4_mm(torch::Tensor& D, torch::Tensor const& A,
#if defined ENABLE_NVFP4 && ENABLE_NVFP4
#if defined ENABLE_NVFP4 && ENABLE_NVFP4
return
cutlass_scaled_fp4_mm_sm100a
(
D
,
A
,
B
,
A_sf
,
B_sf
,
alpha
);
return
cutlass_scaled_fp4_mm_sm100a
(
D
,
A
,
B
,
A_sf
,
B_sf
,
alpha
);
#endif
#endif
TORCH_CHECK_NOT_IMPLEMENTED
(
false
,
"No compiled nvfp4 mm kernel, vLLM should "
TORCH_CHECK_NOT_IMPLEMENTED
(
false
,
"be compiled using CUDA 12.8 and target "
"No compiled nvfp4 mm kernel, vLLM should "
"compute capability 100 or above."
);
"be compiled using CUDA 12.8 and target "
"compute capability 100 or above."
);
}
}
csrc/quantization/fp4/nvfp4_scaled_mm_kernels.cu
View file @
82e0d601
...
@@ -194,8 +194,9 @@ void runGemm(at::Tensor& D, at::Tensor const& A, at::Tensor const& B,
...
@@ -194,8 +194,9 @@ void runGemm(at::Tensor& D, at::Tensor const& A, at::Tensor const& B,
at
::
Tensor
const
&
A_sf
,
at
::
Tensor
const
&
B_sf
,
at
::
Tensor
const
&
A_sf
,
at
::
Tensor
const
&
B_sf
,
at
::
Tensor
const
&
alpha
,
int64_t
m
,
int64_t
n
,
int64_t
k
,
at
::
Tensor
const
&
alpha
,
int64_t
m
,
int64_t
n
,
int64_t
k
,
cudaStream_t
stream
)
{
cudaStream_t
stream
)
{
TORCH_CHECK
(
false
,
"Unsupported CUTLASS version. Set VLLM_CUTLASS_SRC_DIR to "
TORCH_CHECK
(
false
,
"a CUTLASS 3.8 source directory to enable support."
);
"Unsupported CUTLASS version. Set VLLM_CUTLASS_SRC_DIR to "
"a CUTLASS 3.8 source directory to enable support."
);
}
}
#endif // defined(CUTLASS_ARCH_MMA_SM100_SUPPORTED)
#endif // defined(CUTLASS_ARCH_MMA_SM100_SUPPORTED)
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment