LLM-Inference/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-08-12 22:31:11 +04:00
Code Issues Packages Projects Releases Wiki Activity

llama: move suppress_tokens handling to common/sampling (#26276)

Browse Source 
* llama: move suppress_tokens handling to common/sampling

* address security issues

* rm has_logit_bias
This commit is contained in:
Xuan-Son Nguyen
2026-07-29 18:02:30 +02:00
committed by GitHub
parent caa596ab3f
commit afeebe103b
5 changed files with 32 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/llama.h
+3
View File
@@ -1102,6 +1102,9 @@ extern "C" {
LLAMA_API bool llama_vocab_get_add_eos(const struct llama_vocab * vocab);
LLAMA_API bool llama_vocab_get_add_sep(const struct llama_vocab * vocab);
// model-specific suppress tokens (gguf key: tokenizer.ggml.suppress_tokens)
LLAMA_API const llama_token * llama_vocab_get_suppress_tokens(const struct llama_vocab * vocab, int32_t * n_suppress_tokens);
LLAMA_API llama_token llama_vocab_fim_pre(const struct llama_vocab * vocab);
LLAMA_API llama_token llama_vocab_fim_suf(const struct llama_vocab * vocab);
LLAMA_API llama_token llama_vocab_fim_mid(const struct llama_vocab * vocab);