Skip to content

Commit

Permalink
Ditch debug exception
Browse files Browse the repository at this point in the history
  • Loading branch information
ZeronSix committed Feb 26, 2025
1 parent 34e6b8f commit a24ba57
Showing 1 changed file with 0 additions and 14 deletions.
14 changes: 0 additions & 14 deletions inference_lib/src/aqlm/inference_kernels/cuda_kernel.cu
Original file line number Diff line number Diff line change
Expand Up @@ -6,8 +6,6 @@

#include <iostream>

#include <stdexcept>

template<bool use_bfloat16, size_t group_size>
__global__ void Code1x16MatVec(
const int4* __restrict__ A,
Expand Down Expand Up @@ -609,9 +607,6 @@ void code2x8_matvec_cuda(
cudaFuncSetAttribute(
CodeKx8MatVec<use_bfloat16, 2>, cudaFuncAttributeMaxDynamicSharedMemorySize, shared
);
if (cudaGetLastError() != cudaSuccess) {
throw std::runtime_error("618");
}
CodeKx8MatVec<use_bfloat16, 2><<<blocks, threads, shared, stream>>>(
(const int4*) A,
(const int4*) B,
Expand All @@ -620,9 +615,6 @@ void code2x8_matvec_cuda(
prob_m,
prob_k
);
if (cudaGetLastError() != cudaSuccess) {
throw std::runtime_error("628");
}
}
}

Expand Down Expand Up @@ -701,9 +693,6 @@ void code2x8_dequant_cuda(
prob_m,
prob_k
);
if (cudaGetLastError() != cudaSuccess) {
throw std::runtime_error("722");
}
} else {
cudaFuncSetAttribute(
CodeKx8Dequant<false, 2>, cudaFuncAttributeMaxDynamicSharedMemorySize, shared
Expand All @@ -715,9 +704,6 @@ void code2x8_dequant_cuda(
prob_m,
prob_k
);
if (cudaGetLastError() != cudaSuccess) {
throw std::runtime_error("748");
}
}
}
}

0 comments on commit a24ba57

Please sign in to comment.