]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2026-08-12 Georgi Gerganovsync : ggml
2026-08-12 Daniel Beveniuscmake : add config version support (ggml/1582)
2026-08-12 Chipmunkserver : support slot save/restore with media inputs...
2026-08-12 parabelboiui: add read_media tool (#25877)
2026-08-12 Hongqiang Wangopencl: default FA c8 cluster width to 16 on X1E (...
2026-08-12 Georgi Gerganovtests : update speculative params (#26925)
2026-08-12 michaeltrabalka... vulkan: add TQ2_0 (ternary) support (#25850)
2026-08-12 Oğuzhan Akkayawavtokenizer-dec : bound posnet/convnext block_count...
2026-08-12 Wang Zhiyuconvert : handle per_layer_config in Gemma4 (transforme...
2026-08-12 lhezopencl: use flat mv q5_k when weight exceeds image1d_bu...
2026-08-11 ruanslvchat : fix muse-glimmer detection of tool calls after...
2026-08-11 Sigbjørn Skjæretci : add missing release check (#26923)
2026-08-11 Rafail GiavrimisCUDA: only disable CUDA graphs when mul_mat_id actually...
2026-08-11 0cuda : add warp-per-row wkv7 kernel for single-token...
2026-08-11 Georgi Gerganovspec : update speculative-simple (#26904)
2026-08-11 Aldehir Rojaschat : tighten bare function parsing for Qwen models...
2026-08-11 Sigbjørn Skjæretci : add windows-rocm to check-release (#26897)
2026-08-11 Bartowskiimatrix.cpp: Move finite check and only check touched...
2026-08-11 Niklas Wenzelrequirements: use stable torch packages on s390x (...
2026-08-11 ynankaniconvert : keep quantization scales for nemotron --mtp...
2026-08-11 lnigamDflash support for nemotron-3.5 (#26905)
2026-08-11 Xuan-Son Nguyenmtmd: support pocket-tts (#26871)
2026-08-11 Tom Tanui: fix context gauge for single-model usage (#25738)
2026-08-11 uvosci: hip-quality-check: update vgpr spill ignore list...
2026-08-11 Daniel Beveniusmodel-conversion : use save_output_data for causual...
2026-08-11 Georgi Gerganovtests : fix running server tests on windows (#26889)
2026-08-11 Ruben Ortlamllama: add default load-mode auto, which avoids mmap...
2026-08-11 Georgi Gerganovtests : clean-up server test, use `tests.sh` in ci...
2026-08-11 Jim Wutests : disable backend sampler hip multi output (...
2026-08-11 Junmo Kimmodel : fix SWA not being enabled for EXAONE 4.5 (...
2026-08-11 Aldehir Rojascommon/peg : suppress incomplete escape sequences ...
2026-08-11 Masashi Yoshimuraggml-webgpu: fix CI errors from #25025 and #25262 ...
2026-08-10 Gaurav GargAddress review comment of PR 25532 (#26852)
2026-08-10 Hongqiang Wangopencl: transpose the K tile in local memory for FA...
2026-08-10 Mario Limoncielloci : target ROCm 7.14 for build and release (#25775)
2026-08-10 Gaurav Gargllama : support multi-output backend sampling (#25532)
2026-08-10 Hitesh Chopraggml-cpu : fix CPU affinity mask being ignored on Andro...
2026-08-10 Yash Raj Pandeyggml : require contiguous src for ROLL on CUDA and...
2026-08-10 Pascalui: UI/chat form follow ups (#26743)
2026-08-10 Sigbjørn Skjæretci : don't specify python version in server-sanitize...
2026-08-10 Pascalserver: add more tool isolation support (ssh remote...
2026-08-10 Pedro Cuencamodel: Muse Glimmer Support (#26841)
2026-08-10 Guido Imperialechat : Align Laguna-S-2.1 chat template to huggingface...
2026-08-10 Pascalvendor: sync subprocess.h and drop local patches (...
2026-08-10 Pedro Cuencallama: Restore quantization of mmprojs (#26818)
2026-08-10 shivamkumard... ci: Add support for CUDA 13.4 ARM64 builds for Windows...
2026-08-10 Ruixiang Wangmodel: add MTP support for Nemotron model (#26725)
2026-08-10 Alessandro... vendor : update cpp-httplib to 0.53.0 (#26821)
2026-08-10 Bar Haimmodel : Granite-Switch Architecture (#25107)
2026-08-10 Georgi Gerganovreadme : remove dev branches (#26832)
2026-08-10 Aleksander... ui: Linting & Formatting scripts (#26819)
2026-08-10 Pascalserver: gate the docker tools runtime tests on a real...
2026-08-10 Caleb DeLeeuwmodel-saver : fix expert shared/chunk FFN length key...
2026-08-10 Eveci: fix the ctest sanitize runs (#26593)
2026-08-10 Masashi Yoshimuraggml-webgpu : refactor several wgsl files and simplify...
2026-08-09 Pascalui: degrade the working directory picker when file...
2026-08-09 Xuan-Son Nguyenci: add pr-draft-label (#26801)
2026-08-09 Hao-Chen2337ggml-cpu : fix missing Q5_0 dispatch in SpaceMiT backen...
2026-08-09 Aaron Teoci: rm `GGML_HIP_ROCWMMA_FATTN` (#26760)
2026-08-08 Pascalserver: report the isolate working directory from get_i...
2026-08-08 Rafail GiavrimisCUDA: fuse rms_norm + mul + rope (+ view + set_rows...
2026-08-08 Pascalserver, ui: only offer a working directory when a tool...
2026-08-08 Xuan-Son Nguyenserver: add initial tool isolation support (via docker...
2026-08-08 Rafail GiavrimisCUDA: fix thread/block count in quantized cpy kernel...
2026-08-07 Pascaltts: account for the vocoder pass in the timings line...
2026-08-07 Aleksander... allozaur/feat/chat form contenteditable (#26717)
2026-08-07 Georgi Gerganovtests : speed-up server test suite 3x (#26734)
2026-08-07 Aleksander... allozaur/feat/chat slash commands (#26716)
2026-08-07 Titaniumtownsycl: coalesce the ssm_conv window loads (#26612)
2026-08-07 robertomeronimetal : fix NORM/RMS_NORM for row lengths that leave...
2026-08-07 Aleksander... ui: Filesystem `@mentions` for Chat Form (#26715)
2026-08-07 Xuan-Son Nguyenmtmd: fix longest_edge ignoring min/max pixels (#26638)
2026-08-07 Georgi Gerganovsync : ggml
2026-08-07 Georgi Gerganovggml : bump version to 0.19.0 (ggml/1581)
2026-08-07 Daniel Beveniusserver : clarify comment in eval_llama_cmpl_schema...
2026-08-07 Emanuil Rusevwebui: load the model selected via ?model= when ?load...
2026-08-07 Niklas Wenzelui: set npm `min-release-age` to protect against supply...
2026-08-07 Xuan-Son Nguyenserver: (router) add LRU scheduler (#26572)
2026-08-07 Xuan-Son Nguyenserver: (router) do not evict busy models (#26567)
2026-08-07 Pascalmtmd: stop feeding the text stream again during Qwen3...
2026-08-07 Kilian Huggml : add aarch64 HWCAP fallbacks and fix fp16 variant...
2026-08-07 Pascalui: read model modalities from the router model list...
2026-08-07 Masato NakasakaMitigate crashing issue on Windows MSYS2 UCRT64 environ...
2026-08-07 Chris Leesycl: fix UE4M3 parsing (#25608)
2026-08-07 Titaniumtownsycl: *glu flat path (#26354)
2026-08-07 Neo Zhangsycl : Support DSv4 OPs: LIGHTNING_INDEXER,DSV4_HC_COMB...
2026-08-07 Neo Zhangsycl : update guide Q&A and script for device setting...
2026-08-07 Neo Zhangsycl : fix error Error OP FLASH_ATTN_EXT on arc770...
2026-08-07 Neo Zhangsycl : enhance OP set_rows to support all missed data...
2026-08-07 David Friehscuda: fix warnings for unused variable/function (#26688)
2026-08-07 Niklas Wenzelci: abort if build requirements are missing (#26368)
2026-08-07 JamePengmetal : avoid `threadgroup` matrix array instantiation...
2026-08-06 Xuan-Son Nguyenmtmd: add chunk save/load function (#26645)
2026-08-06 Xuan-Son Nguyenserver: fix empty response for /cors-proxy (#26656)
2026-08-06 Sigbjørn Skjæretconvert : fix DeepseekV4 rope parameters with transform...
2026-08-06 Georgi Gerganovmodel-loader : fix quantized reshaped tensor strides...
2026-08-06 Csaba Kecskemeticonvert : accept "ExaoneMoeForCausalLM" arch spelling...
2026-08-06 Jim Wuci : onboard AMD ROCm CI with gfx1151 fixes (#26544)
2026-08-06 Daniel Beveniusmodel-conversion : add --model-name to conversion scrip...
2026-08-06 Ruben Ortlamvulkan: fix submission batching size, add debug tools...
next