tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-08-25 09:38:35 -04:00
Code Issues Packages Projects Releases Wiki Activity

speculative : update default params (#11954)

Browse Source 
* speculative : update default params

* speculative : do not discard the last drafted token
This commit is contained in:
Georgi Gerganov
2025-02-19 13:29:42 +02:00
committed by GitHub
parent 9626d9351a
commit abd4d0bc4f
4 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
common/speculative.h
View File

@@ -9,7 +9,7 @@ struct common_speculative_params {
int n_draft = 16; // max drafted tokens
int n_reuse = 256;
float p_min = 0.9f; // min probability required to accept a token in the draft
float p_min = 0.75f; // min probability required to accept a token in the draft
};
struct common_speculative * common_speculative_init(struct llama_context * ctx_dft);