char-fork/llama.cpp

git clone https://git.t4t.associates/char-fork/llama.cpp

Anav PrasadCUDA: fuse SSM_CONV + ADD(bias) + SILU (#22478)098705a29

char/aliceai-t5-moe
176 B3 linesraw
1#include "common.cuh"
2
3void ggml_cuda_op_ssm_conv(ggml_backend_cuda_context & ctx, ggml_tensor * dst, ggml_tensor * bias_add_node = nullptr, ggml_tensor * silu_dst = nullptr);