ollama/llama/ggml-cuda/concat.cuh
2024-09-03 21:15:12 -04:00

6 lines
137 B
Plaintext
Vendored

#include "common.cuh"
#define CUDA_CONCAT_BLOCK_SIZE 256
void ggml_cuda_op_concat(ggml_backend_cuda_context & ctx, ggml_tensor * dst);