Pull Requests ggml-org/llama.cpp

Use fp32 in cuBLAS V100 to avoid overflows, env variables to override cuBLAS compute type Nvidia GPU ggml

#19959 opened 2026-02-27 19:25 by wallentri88

llama : add native param2moe architecture support model python

#19958 opened 2026-02-27 19:01 by iambhuvan

tools : enable kvu in perplexity for hellaswag, winogrande, multiple-choice examples

#19954 opened 2026-02-27 14:37 by angt

scripts : improve get-wikitext-2.sh script

#19952 opened 2026-02-27 13:19 by angt

[New quant] Q3_PT examples python ggml

#19941 opened 2026-02-26 22:15 by pwilkin

webui: use date in more human readable exported filename examples server

#19939 opened 2026-02-26 20:15 by woof-dog

scripts: ini_to_opencode.py script python

#19938 opened 2026-02-26 18:13 by am17an

fix dots.ocr: correct RoPE sections and FFN tensor mapping examples python

#19936 opened 2026-02-26 16:27 by anthony-maio

common : update completion executables list [no ci]

#19934 opened 2026-02-26 14:12 by danbev

tool parser: add GigaChatV3/3.1 models support in PEG format testing

#19931 opened 2026-02-26 13:07 by Mishusha

metal: add CONV_3D ggml Apple Metal

#19927 opened 2026-02-26 11:24 by Ra5hidIslam

llama/ggml: multi-GPU pipeline parallelism (xdev host staging) + faster model loading Nvidia GPU ggml

#19922 opened 2026-02-26 09:43 by mxxm-t

ggml-cuda: add mem check for fusion Nvidia GPU ggml

#19916 opened 2026-02-26 05:53 by am17an

vendors: update miniaudio library to 0.11.24 script python

#19914 opened 2026-02-26 04:37 by data-man

test-backend-ops: allow loading tests from JSON and parsing model operators into JSON testing examples

#19896 opened 2026-02-25 15:13 by 0cc4m

[ggml-quants] Add memsets and other fixes for IQ quants ggml

#19861 opened 2026-02-24 20:04 by bartowski1182

server : add default-model preset and fallback logic examples server

#19855 opened 2026-02-24 16:30 by mikhail-shevtsov-wiregate

ggml-webgpu: Support non-contiguous `src0` and overlapping `src0/src1` in binary ops testing ggml

#19850 opened 2026-02-24 12:20 by yomaytk

server : add chat truncation to keep chat going testing examples python server

#19841 opened 2026-02-23 21:36 by ltoniazzi

opencl: add optimized q4_1 mm kernel for adreno ggml OpenCL

#19840 opened 2026-02-23 21:36 by shaofeiqi

sampling : support multiple outputs per sequence testing examples server

#19833 opened 2026-02-23 13:51 by danbev

Add Aya 101 multi-lingual translation support to llama.cpp examples python server

#19832 opened 2026-02-23 13:23 by Acceldium

common : refactor cache to use hierarchical directory layout

#19828 opened 2026-02-23 12:30 by angt

Kimi Linear block implementation model

#19827 opened 2026-02-23 12:14 by ymcki

implemented max pooling for embeddings examples python server

#19812 opened 2026-02-22 19:03 by lorenzocesconetto

llama: end-to-end tests model testing

#19802 opened 2026-02-22 10:59 by JohannesGaessler

Add model metadata loading from huggingface for use with tests requiring real model data testing

#19796 opened 2026-02-22 05:07 by bartowski1182

tools : add learning-cache tool for persistent latent context examples

#19791 opened 2026-02-22 00:55 by arkavo-com

[WIP] ggml-hexagon: convert f32 to f16 - fa opt part4 ggml

#19780 opened 2026-02-21 15:13 by chraac

Clean up per-thread parameter buffer pool and job submission logic ggml

#19772 opened 2026-02-20 23:27 by nikhilJain17