]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2026-06-04 Andreifix(mtmd): handle Gemma 4 audio projector embedding...
2026-06-04 Georgi Gerganovreadme : add status badges (#24104)
2026-06-04 Georgi Gerganovtests : refactor test-save-load-state to accept token...
2026-06-04 Georgi Gerganovmetal : reduce rset heartbeat from 500ms -> 5ms (#24074)
2026-06-04 Reese Levineggml-webgpu: FlashAttention refactor + standardize...
2026-06-04 rehan-10xengineerggml-cpu: extend RVV quantization vec dot to higher...
2026-06-04 Todd Malsbarysycl : Improve SYCL doc (#23025)
2026-06-03 Andreimtmd: fix Gemma 4 unified FPE (#24088)
2026-06-03 Aman Guptaqwen35: use post-norm hidden state for MTP (#24025)
2026-06-03 Xuan-Son Nguyenmtmd: enable non-causal vision for gemma 4 unified...
2026-06-03 Xuan-Son Nguyenmtmd, model: allow skip build_vit() (#24077)
2026-06-03 Aleksander... ui: Mermaid Diagrams in chat + interactive preview...
2026-06-03 Andreas KieslingerAvoid PDL race conditions by disabling __restrict__...
2026-06-03 Charles Xuggml-cpu: use runtime SVE width in FWHT (#24059)
2026-06-03 Aman Guptacuda: reserve space for quantize kv-cache at startup...
2026-06-03 Georgi Gerganovtests : add support for qwen3 SSM archs (#24031)
2026-06-03 Alessandro... update BoringSSL to 0.20260526.0 (#23794)
2026-06-03 Georgi Gerganovci : disable ccache for msvc windows release jobs ...
2026-06-03 Ryan Mangenoarg : removed unecesary mmproj download when users...
2026-06-02 lhezopencl: use flat variants of q4_K and q6_K gemv for...
2026-06-02 Max Krasnyanskyhexagon: profiler output fix and script updates (#24042)
2026-06-02 Mikhail Podvitskiimodel: add Mellum architecture (#23966)
2026-06-02 Hans Florianmodel : support granite multilingual embeddings R2...
2026-06-02 Piotr Wilkin... StepFun 3.5 MTP (#23274)
2026-06-02 Daniel Beveniuscommon : fix state save in common_prompt_batch_decode...
2026-06-02 Xuan-Son Nguyenserver: add SSE ping interval (#24013)
2026-06-02 Georgi Gerganovci : reduce self-hosted server workflow jobs (#24012)
2026-06-02 Mikhail Podvitskiidocs : update HOWTO-add-model.md (#23883)
2026-06-02 Marcos Del... ui: simplify network error handling (#23431)
2026-06-02 Aleksander... ui: Add Thinking mode toggle with reasoning effort...
2026-06-02 Georgi Gerganovkv-cache : SWA checkpoints store only non-masked cells...
2026-06-02 forforever73convert : support Step3.7-Flash (#23845)
2026-06-02 Georgi Gerganovllama : deprecate `llama_set_warmup` (#24009)
2026-06-02 Max Krasnyanskyhexagon: MUL_MAT, MUL_MAT_ID, FLASH_ATTN and GDN cleanu...
2026-06-02 Todor Boinovskihexagon: add gelu_quick (#24007)
2026-06-02 Pascalserver: real-time reasoning interruption via control...
2026-06-02 Anav Prasadclean up unused variables warnings (#23975)
2026-06-02 lhezopencl: fix compiler warnings for non-adreno path ...
2026-06-01 Masashi Yoshimurarevert to using global_invocation_id for cpy shader...
2026-06-01 Georgi Gerganovspeculative : fix n_outputs_max and remove draft-simple...
2026-06-01 Christian Hoener... nix : add nix-nodejs facilities to build Web UI (#23846)
2026-06-01 shaofeiqiopencl: add basic support for q5_0 and q5_1 (#23548)
2026-06-01 Adrien Gallouëtvendor : update cpp-httplib to 0.46.1 (#23980)
2026-06-01 Aman Guptallama: limit max outputs of `llama_context` (#23861)
2026-06-01 Shrivas Shankarmetal: template GLU kernels to support f16/f32 (#23882)
2026-06-01 Jeff Bolzvulkan: don't hold the device mutex while compiling...
2026-06-01 Winston Mavulkan: reduce host memory lock contention (#23376)
2026-06-01 o7sivocab: add normalizer.lowercase support to WPM (#23899)
2026-06-01 Johannes GäßlerTP: quantized KV cache support (#23792)
2026-06-01 Georgi Gerganovsecurity : disable private disclosures (#23963)
2026-06-01 Junwon Hwangmodel: Add EXAONE 4.5 implementations (#21733)
2026-06-01 Matt Corallovulkan: Block-load Q3_K/Q6_K block data and subtract...
2026-06-01 Winston Mavulkan: Removed unused functions (#23175)
2026-06-01 Aldehir Rojascommon : support manually triggering the reasoning...
2026-06-01 Georgi Gerganovci : add missing Linux label to cpu-x64-high-perf runne...
2026-06-01 Neo Zhang[SYCL] Support Q4_1, Q5_0, Q5_1 in Flash-attention...
2026-06-01 Neo Zhang[SYCL] Add more types in GET_ROWS OP (#23710)
2026-06-01 Neo Zhangsycl : Optimize Q3_K mul_mat by reorder (#23725)
2026-06-01 Eveci: remove redundant or duplicate jobs (#23927)
2026-05-31 Eric Zhangserver : handle If-None-Match weak ETags (#23916)
2026-05-31 Georgi Gerganovci : limit trigger paths for the CPU workflow (#23938)
2026-05-31 o7sivocab : add tokenizer support for jina-embeddings-v2...
2026-05-31 Eric Zhangui: fix ETag truncation with MSVC compiler (#23917)
2026-05-31 Vladislavdocs : update ZenDNN docs for Q8 support (#23791)
2026-05-31 Ruben Ortlamllama: only use one iGPU device by default (#23897)
2026-05-30 Pascalwebui: add custom CSS injection via config (#23904)
2026-05-30 Gaurav GargSupport `-fa auto` in llama-bench (#23714)
2026-05-30 lhezopencl: support bf16 by converting to f16 (#23839)
2026-05-30 Pascalui: exclude generated build dirs from prettier and...
2026-05-30 Johannes GäßlerTP: fix granularity for Qwen 3.5/3.6 + 3 GPUs (#23843)
2026-05-30 Georgi Gerganovmetal : restore im2col implementation for large kernels...
2026-05-30 Xuan-Son Nguyentest: (test-llama-archs) log the config name first...
2026-05-30 Georgi Gerganovci : update ios-xcode release job to macos-26 (#23906)
2026-05-30 Jinyang Heggml : add some lsx support (#23798)
2026-05-30 Ruben Ortlamvulkan: add Flash Attention support for BFloat16 KV...
2026-05-30 Georgi Gerganovci : fix s390x release job (#23898)
2026-05-30 Georgi Gerganovci : clear cache instead of "no timestamp" keys + fix...
2026-05-30 Radoslav Gerganovllama : do not skip iGPU when only RPC devices are...
2026-05-29 Xuan-Son Nguyenserver: in SSE mode, send HTTP headers when slot starts...
2026-05-29 Reese Levineggml-webgpu: Check earlier for WebGPU required features...
2026-05-29 Reese Levineggml-webgpu: add q4_0/q8_0 SET_ROWS (#23760)
2026-05-29 Ruixiang Wangserver-bench : add speed-bench for speculative decoding...
2026-05-29 Pascalapp: add llama update self updater (#23865)
2026-05-29 ValdikSSui: handle audio/vnd.wave as audio WAV file (#23754)
2026-05-29 Tarek Dakhranvocab : support tokenizer for LFM2.5-8B-A1B (#23826)
2026-05-29 Sigbjørn Skjæretgraph : ensure DS32 kq_mask_lid is F32 (#23864)
2026-05-29 Xuan-Son Nguyenserver: remove obsolete scripts (#23870)
2026-05-29 Georgi Gerganovci : update macos release to use macos-26 runner (...
2026-05-29 Xuan-Son Nguyendownload: add option to skip_download (#23059)
2026-05-29 Saba Fallahmtmd: Add DeepSeekOCR 2 Support (#20975)
2026-05-29 Oliver SimonsCUDA: Check PTX version on host side to guard PDL dispa...
2026-05-29 Xuan-Son Nguyenserver: bump timeout to 3600s (#23842)
2026-05-29 fairydreamingmodel : support for DeepseekV32ForCausalLM with generic...
2026-05-29 Aman Guptallama: use f16 mask for FA to save VRAM (#23764)
2026-05-29 Georgi Gerganovsync : ggml
2026-05-29 Georgi Gerganovggml : bump version to 0.13.1 (ggml/1523)
2026-05-29 Omid Azizingram-mod : Add missing include (#23857)
2026-05-29 Aman Guptallama: add llm_graph_input_mtp (#23643)
2026-05-29 Adrien Gallouëtapp : move licences to llama-app (#23824)
2026-05-29 Andreas Kieslingercuda : disables launch_fattn PDL enrollment due to...
next