[quant kernel] sgl-kernel support per_tensor_quant fp8 (#3786)

This commit is contained in:
Xiaoyu Zhang
2025-03-07 10:05:43 +08:00
committed by GitHub
parent 361971b859
commit ad55f17182
8 changed files with 343 additions and 0 deletions

View File

@@ -106,6 +106,7 @@ sources = [
"src/sgl-kernel/csrc/gemm/fp8_blockwise_gemm_kernel.cu",
"src/sgl-kernel/csrc/gemm/int8_gemm_kernel.cu",
"src/sgl-kernel/csrc/gemm/per_token_group_quant_fp8.cu",
"src/sgl-kernel/csrc/gemm/per_tensor_quant_fp8.cu",
"src/sgl-kernel/csrc/moe/moe_align_kernel.cu",
"src/sgl-kernel/csrc/speculative/eagle_utils.cu",
"src/sgl-kernel/csrc/speculative/speculative_sampling.cu",