Pull Requests ggerganov/llama.cpp

server : fix division by zero when reporting stats examples server

#16501 by ggerganov was merged 2025-10-10 19:15

vocab : mark EOT token for Granite models

#16499 by ggerganov was merged 2025-10-10 14:17

server : log requests to /v1/completions examples server

#16495 by rgerganov was merged 2025-10-10 10:22

[feat]: test

#16493 by JingliangGao was closed 2025-10-10 08:04

webui: updated the chat service to only include max_tokens in the req… examples server

#16489 by ServeurpersoCom was merged 2025-10-09 20:54

server : return HTTP 400 if prompt exceeds context length examples python server

#16486 by rgerganov was merged 2025-10-10 14:11

No markdown in cot examples server

#16483 by ServeurpersoCom was merged 2025-10-09 15:36

server : add option to debug the slot contents examples server

#16482 by ggerganov was merged 2025-10-09 14:18

Dont define _XOPEN_SOURCE on AIX ggml

#16481 by mehendarkarprajwal was merged 2025-10-10 08:15

[SYCL] refactor soft_max, add soft_max_back ggml SYCL

#16472 by NeoZhangJianyu was merged 2025-10-09 07:25

CUDA Copy Kernel for Contiguous Tensors for GGML CPY OP Nvidia GPU ggml

#16471 by anavp-nvidia was closed 2025-10-09 12:27

Minor set_rows Optimization ggml

#16468 by neha-ha was closed 2025-10-08 02:36

server : Fixed canceling pending tasks examples server

#16467 by issixx was merged 2025-10-08 08:20

llama : support LiquidAI LFM2-MoE hybrid model model python hot

#16464 by tdakhran was merged 2025-10-07 18:03

ci: add ARM64 Kleidiai build and test support devops

#16462 by sudhiarm was merged 2025-10-09 08:13

server : add `/v1/health` endpoint examples server

#16461 by ggerganov was merged 2025-10-07 12:57

kleidiai: kernel interface refactoring ggml

#16460 by chaxu01 was merged 2025-10-09 07:29

kleidiai: kernel interface refactoring ggml

#16459 by chaxu01 was closed 2025-10-07 10:19

presets : fix pooling param for embedding models

#16455 by ggerganov was merged 2025-10-07 07:32

common : increase default `n_ctx_checkpoints` from 3 to 8

#16453 by ddh0 was closed 2025-10-08 17:00

ggml webgpu: profiling, CI updates, reworking of command submission devops ggml

#16452 by reeselevine was merged 2025-10-07 20:48

metal : various optimizations + refactoring ggml Apple Metal

#16446 by ggerganov was merged 2025-10-07 05:21

ci : remove missing reranker model files devops

#16444 by danbev was merged 2025-10-06 12:56

ggml-cpu : fix leftover handling in ggml_vec_scale_f32 for SVE ggml

#16443 by danbev was merged 2025-10-06 12:17

memory : use sequential equal splits for recurrent modules

#16442 by ggerganov was merged 2025-10-07 05:24

rpc : update documentation examples

#16441 by rgerganov was merged 2025-10-07 06:59

server : improve context checkpoint logic examples server

#16440 by ggerganov was merged 2025-10-08 07:57

Granite Docling stopping testing examples

#16438 by gabe-l-hart was merged 2025-10-06 16:59

server: update readme to mention n_past_max metric examples server

#16436 by okuvshynov was merged 2025-10-06 07:53

rpc : check src buffer when copying tensor ggml

#16421 by rgerganov was merged 2025-10-04 13:22