mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-09-02 13:24:28 -04:00
432df2d5f9
* RoPE: fix back, CUDA support for back + noncont. * fix comments reg. non-cont. RoPE support [no-ci]
8 lines
215 B
Plaintext
8 lines
215 B
Plaintext
#include "common.cuh"
|
|
|
|
#define CUDA_ROPE_BLOCK_SIZE 256
|
|
|
|
void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
|
|
|
|
void ggml_cuda_op_rope_back(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
|