llama-cpp-sys-4 0.7.0

Low Level Bindings to llama.cpp
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
#ifndef GGML_SYCL_FATTN_ONEDNN_HPP
#define GGML_SYCL_FATTN_ONEDNN_HPP

#include "common.hpp"

// Static-only check: fused-XMX oneDNN Graph SDPA path==flash-attn op
// (f16 KV, no softcap/ALiBi, single stream, tuned head_dim, prefill-sized q.)
bool ggml_sycl_flash_attn_ext_onednn_supported(const ggml_tensor * dst, bool use_shape_limit = true);

// True when the oneDNN path binds an F16 KV cache in place instead of staging a dense copy of
// it. Depends only on the types and strides of K and V, so the answer holds for every call.
bool ggml_sycl_fattn_onednn_binds_kv(const ggml_tensor * K, const ggml_tensor * V);

// Run flash attention through oneDNN's fused xmx SDPA
// execute the cached SDPA partition, write the f32 dst. Falls back to the TILE kernel on any failure.
void ggml_sycl_flash_attn_ext_onednn(ggml_backend_sycl_context & ctx, ggml_tensor * dst);

#endif // GGML_SYCL_FATTN_ONEDNN_HPP