tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-07-17 08:14:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

CUDA: fix scratch malloced on non-main device (#3220)

Browse Source
This commit is contained in:
Johannes Gäßler
2023-09-17 14:16:22 +02:00
committed by GitHub
parent b541b4f0b1
commit 578d8c8f5c
1 changed files with 1 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
ggml-cuda.cu
View File

@ -6970,6 +6970,7 @@ void ggml_cuda_assign_scratch_offset(struct ggml_tensor * tensor, size_t offset)
return;
}
if (g_scratch_buffer == nullptr) {
ggml_cuda_set_device(g_main_device);
CUDA_CHECK(cudaMalloc(&g_scratch_buffer, g_scratch_size));
}