root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-13 04:00:16 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

Disable flash attention for Gemma2

Browse Source
This commit is contained in:
Andrei Betlen 2024-06-28 16:00:20 -04:00
parent d1137c20f1
commit f4424c150f
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/llama.cpp
View File

@ -17402,6 +17402,11 @@ struct llama_context * llama_new_context_with_model(
params.flash_attn = false;
}
if (params.flash_attn && model->arch == LLM_ARCH_GEMMA2) {
LLAMA_LOG_WARN("%s: flash_attn is not compatible with Gemma2 - forcing off\n", __func__);
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;