Public/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-19 09:15:04 +02:00
Code Issues Packages Projects Releases Wiki Activity

sampling : add min-p backend sampler

Browse Source
This commit is contained in:
Daniel Bevenius
2025-11-26 10:50:58 +01:00
parent f23b306cc5
commit b45d504e70
4 changed files with 188 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/llama.h
+3
View File
@@ -1405,6 +1405,9 @@ extern "C" {
/// @details Distribution sampling on backend - final sampling step that selects a token
LLAMA_API struct llama_sampler * llama_sampler_backend_init_dist(uint32_t seed);
/// @details Min-P filtering on backend - filter tokens with a probability less than p times the maximum probability.
LLAMA_API struct llama_sampler * llama_sampler_backend_init_min_p(float p);
// Returns the seed used by the sampler if applicable, LLAMA_DEFAULT_SEED otherwise
LLAMA_API uint32_t llama_sampler_get_seed(const struct llama_sampler * smpl);