fix: compilation of gptq_marlin_gemm object (#800)

* fix: compilation of gptq_marlin_gemm object * add devlop to codespell ignore
PygmalionAI · Nov 2, 2024 · 5b0eabe · 5b0eabe
1 parent 7e2d2e7
commit 5b0eabe
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 7 deletions.
diff --git a/kernels/torch_bindings.cpp b/kernels/torch_bindings.cpp
@@ -147,12 +147,7 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
   ops.impl("gptq_marlin_24_gemm", torch::kCUDA, &gptq_marlin_24_gemm);
 
   // gptq_marlin Optimized Quantized GEMM for GPTQ.
-  ops.def(
-      "gptq_marlin_gemm(Tensor a, Tensor b_q_weight, Tensor b_scales, "
-      "Tensor b_zeros, Tensor g_idx, Tensor perm, Tensor workspace, "
-      "int b_q_type, "
-      "SymInt size_m, SymInt size_n, SymInt size_k, bool is_k_full, "
-      "bool has_zp, bool use_fp32_reduce, bool is_zp_float) -> Tensor");
+  ops.def("gptq_marlin_gemm", &gptq_marlin_gemm);
   ops.impl("gptq_marlin_gemm", torch::kCUDA, &gptq_marlin_gemm);
 
   // gptq_marlin repack from GPTQ.

diff --git a/pyproject.toml b/pyproject.toml
@@ -46,7 +46,7 @@ ignore = [
 ]
 
 [tool.codespell]
-ignore-words-list = "dout, te, indicies, ist, subtile, wit, whit, beseige"
+ignore-words-list = "dout, te, indicies, ist, subtile, wit, whit, beseige, devlop"
 skip = "./tests/,./aphrodite/endpoints/kobold/klite.embd,./kernels/,./tests/benchmarks/sonnet.txt,./docs/"
 
 [tool.isort]