chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

[Kernel Slimming] Migrate NVFP4 kernels to JIT (#19437)

Browse Source
This commit is contained in:
Mohammad Miadh Angkad
2026-03-05 15:22:28 +08:00
committed by GitHub
parent 1bbfed0539
commit 2bdd89a6cd
27 changed files with 2458 additions and 989 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
sgl-kernel/python/sgl_kernel/__init__.py
View File

@@ -51,7 +51,6 @@ from sgl_kernel.gemm import (
int8_scaled_mm,
qserve_w4a8_per_chn_gemm,
qserve_w4a8_per_group_gemm,
scaled_fp4_experts_quant,
scaled_fp4_grouped_quant,
scaled_fp4_quant,
sgl_per_tensor_quant_fp8,
@@ -79,7 +78,6 @@ from sgl_kernel.mamba import (
from sgl_kernel.memory import set_kv_buffer_kernel, weak_ref_tensor
from sgl_kernel.moe import (
apply_shuffle_mul_sum,
cutlass_fp4_group_mm,
fp8_blockwise_scaled_grouped_mm,
fused_qk_norm_rope,
kimi_k2_moe_fused_gate,