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

[Kernel] fp4 marlin kernel (#17687)

Browse Source 
Signed-off-by: Jinzhen Lin <linjinzhen@hotmail.com>
This commit is contained in:
Jinzhen Lin
2025-05-11 10:58:49 +08:00
committed by GitHub
parent ca66a1674c
commit d74e5f37bc
21 changed files with 1216 additions and 331 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
csrc/torch_bindings.cpp
View File

@@ -292,8 +292,8 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
// gptq_marlin Optimized Quantized GEMM for GPTQ.
ops.def(
"gptq_marlin_gemm(Tensor a, Tensor? c_or_none, Tensor b_q_weight, "
"Tensor b_scales, Tensor? b_zeros_or_none, Tensor? g_idx_or_none, "
"Tensor? perm_or_none, Tensor workspace, int b_q_type, "
"Tensor b_scales, Tensor? global_scale, Tensor? b_zeros_or_none, Tensor? "
"g_idx_or_none, Tensor? perm_or_none, Tensor workspace, int b_q_type, "
"SymInt size_m, SymInt size_n, SymInt size_k, bool is_k_full, "
"bool use_atomic_add, bool use_fp32_reduce, bool is_zp_float) -> Tensor",
{stride_tag});