biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Refactor] [1/N] Reorganize kernel abstraction directory (#34055)

Browse Source 
Signed-off-by: BadrBasowid <badr.basowid@gmail.com>
Co-authored-by: vllmellm <vllm.ellm@embeddedllm.com>
Co-authored-by: TJian <tunjian.tan@embeddedllm.com>
This commit is contained in:
BadrBasowid
2026-02-24 14:47:22 +08:00
committed by GitHub
parent 1a6cf39dec
commit 6af03f2394
40 changed files with 331 additions and 213 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
tests/compile/passes/test_fusion.py
View File

@@ -26,24 +26,16 @@ from vllm.config import (
PassConfig,
VllmConfig,
)
from vllm.model_executor.layers.layernorm import RMSNorm
from vllm.model_executor.layers.quantization.kernels.scaled_mm.cutlass import (
CutlassFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.flashinfer import (
FlashInferFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.pytorch import (
from vllm.model_executor.kernels.linear import (
ChannelWiseTorchFP8ScaledMMLinearKernel,
CutlassFP8ScaledMMLinearKernel,
FlashInferFP8ScaledMMLinearKernel,
FP8ScaledMMLinearKernel,
PerTensorTorchFP8ScaledMMLinearKernel,
ROCmFP8ScaledMMLinearKernel,
RowWiseTorchFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.rocm import (
ROCmFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.ScaledMMLinearKernel import ( # noqa: E501
FP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.layernorm import RMSNorm
from vllm.model_executor.layers.quantization.utils.quant_utils import (
GroupShape,
QuantKey,

14
tests/compile/passes/test_silu_mul_quant_fusion.py
View File

@@ -26,22 +26,14 @@ from vllm.config import (
VllmConfig,
set_current_vllm_config,
)
from vllm.model_executor.layers.activation import SiluAndMul
from vllm.model_executor.layers.quantization.kernels.scaled_mm.cutlass import (
from vllm.model_executor.kernels.linear import (
CutlassFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.flashinfer import (
FlashInferFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.pytorch import (
FP8ScaledMMLinearKernel,
PerTensorTorchFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.rocm import (
ROCmFP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.quantization.kernels.scaled_mm.ScaledMMLinearKernel import ( # noqa: E501
FP8ScaledMMLinearKernel,
)
from vllm.model_executor.layers.activation import SiluAndMul
from vllm.model_executor.layers.quantization.utils.fp8_utils import W8A8BlockFp8LinearOp
from vllm.model_executor.layers.quantization.utils.quant_utils import (
GroupShape,