CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

metal : improve FA + improve MoE (#12612)

Browse Source 
* ggml : FA with different K, V head sizes (CPU)

ggml-ci

* metal : add FA with HS=192

* metal : extend FA to support different K and V head sizes

ggml-ci

* metal : add FA vector kernels for heads K 192 and V 128

ggml-ci

* ggml : restrict op on other backends to equal head sizes

ggml-ci

* metal : optimize FA-vec kernel

ggml-ci

* metal : FA remove mq registers

* metal : improve MoE mul_mat_id condition

ggml-ci

* metal : fix comments + remove unnecessary addition

ggml-ci

* metal : avoid too much shared memory usage with mul_mat_id

ggml-ci
This commit is contained in:
Georgi Gerganov
2025-03-28 20:21:59 +02:00
committed by GitHub
parent b86f600723
commit b4ae50810e
10 changed files with 913 additions and 706 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
src/llama-context.cpp
View File

@@ -2316,11 +2316,6 @@ llama_context * llama_init_from_model(
params.flash_attn = false;
}
if (params.flash_attn && model->hparams.n_embd_head_k != model->hparams.n_embd_head_v) {
LLAMA_LOG_WARN("%s: flash_attn requires n_embd_head_k == n_embd_head_v - forcing off\n", __func__);
params.flash_attn = false;
}
if (ggml_is_quantized(params.type_v) && !params.flash_attn) {
LLAMA_LOG_ERROR("%s: V cache quantization requires flash_attn\n", __func__);
return nullptr;