ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-28 10:21:48 +00:00
Code Issues Packages Projects Releases Wiki Activity

server: exclude thinking tokens when finding the slot (#1079)

Browse Source 
refactor find slot

enable by default

Fix load prompt

rename variables

Co-authored-by: firecoperana <firecoperana>
This commit is contained in:
firecoperana
2025-12-22 02:46:45 -06:00
committed by GitHub
parent 21fc9322f9
commit 5562605076
8 changed files with 247 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/server/server-common.h
View File

@@ -171,6 +171,7 @@ std::string tokens_to_str(llama_context* ctx, const llama_tokens& tokens);
// format incomplete utf-8 multibyte character for output
std::string tokens_to_output_formatted_string(const llama_context* ctx, const llama_token token);
struct common_prefix {
size_t first = 0;
size_t second = 0;
@@ -389,6 +390,7 @@ public:
size_t get_common_prefix_exact(const server_tokens& b) const;
llama_tokens get_text_tokens_exclude_think(const llama_context* ctx, const thinking_tokens& think_token) const;
common_prefix get_common_prefix(const llama_context* ctx, const server_tokens& b, bool exact = false) const;
// take first n tokens of tokens list a