Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-19 09:15:18 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit 'a61c8bc3bfae4f86b8205535bcea73f476b28c2c' into concedo_experimental

Browse Source 
# Conflicts:
#	ggml/src/ggml-opencl/CMakeLists.txt
#	ggml/src/ggml-opencl/ggml-opencl.cpp
#	ggml/src/ggml-webgpu/ggml-webgpu.cpp
#	scripts/pr2wt.sh
#	src/llama-model.cpp
#	tools/CMakeLists.txt
#	tools/mtmd/CMakeLists.txt
#	tools/mtmd/clip.cpp
#	tools/mtmd/clip.h
This commit is contained in:
Concedo
2026-01-13 23:06:50 +08:00
parent 612c19afe7 a61c8bc3bf
commit 0dc18c668c
35 changed files with 3867 additions and 2025 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/llama.h
+3 -1
View File
@@ -1295,7 +1295,9 @@ extern "C" {
// available samplers:
LLAMA_API struct llama_sampler * llama_sampler_init_greedy(void);
LLAMA_API struct llama_sampler * llama_sampler_init_dist (uint32_t seed);
/// seed == LLAMA_DEFAULT_SEED to use a random seed.
LLAMA_API struct llama_sampler * llama_sampler_init_dist(uint32_t seed);
/// @details Top-K sampling described in academic paper "The Curious Case of Neural Text Degeneration" https://arxiv.org/abs/1904.09751
/// Setting k <= 0 makes this a noop