vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Revert back to int64_t.

Browse source
This commit is contained in:
DAN™ 2024-04-28 18:33:37 -04:00
parent 91c10ef225
commit 0258f9bd3d
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml-cuda.cu
View file

@ -1225,7 +1225,7 @@ static void ggml_cuda_op_mul_mat_cublas(
// the main device has a larger memory buffer to hold the results from all GPUs
// ldc == nrows of the matrix that cuBLAS writes into
int ldc = id == ctx.device ? ne0 : row_diff;
int64_t ldc = id == ctx.device ? ne0 : row_diff;
const int compute_capability = ggml_cuda_info().devices[id].cc;