mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-08-12 19:37:53 -04:00
8 lines
228 B
Plaintext
8 lines
228 B
Plaintext
#include "common.cuh"
|
|
|
|
#define CUDA_SOFT_MAX_BLOCK_SIZE 1024
|
|
|
|
void ggml_cuda_op_soft_max(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
|
|
|
|
void ggml_cuda_op_soft_max_back(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
|