vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ggml : ggml_flash_attn_ext() support ALiBi (CPU)

Browse source
This commit is contained in:
Georgi Gerganov 2024-05-10 11:33:34 +03:00
parent d0592d495d
commit 166e60bf9b
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
3 changed files with 25 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

3
ggml.h
View file

@ -1731,7 +1731,8 @@ extern "C" {
struct ggml_tensor * k,
struct ggml_tensor * v,
struct ggml_tensor * mask,
float scale);
float scale,
float max_bias);
GGML_API void ggml_flash_attn_ext_set_prec(
struct ggml_tensor * a,