]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2025-05-31 Georgi Gerganovllama : deprecate explicit kv_self defrag/update calls...
2025-05-31 Georgi Gerganovllama : use n_swa + n_ubatch cells for SWA cache (...
2025-05-31 igardevwebui : Replace alert and confirm with custom modals...
2025-05-31 Georgi Gerganovllama : auto-batch preparation (#13845)
2025-05-31 Xuan-Son Nguyenmtmd : drop `_shared` from `libmtmd` name, merge helper...
2025-05-31 Georgi Gerganovkv-cache : refactor + add llama_memory_state_i (#13746)
2025-05-31 Shawn yangCUDA: add a prop in ggml_cuda_device_infor for distingu...
2025-05-30 Johannes GäßlerCUDA: fix typo in FlashAttention code (#13926)
2025-05-30 Diego Devesasched : avoid changing cur_copy when a graph is already...
2025-05-30 Georgi Gerganovparallel : increase the variability of the prompt lengt...
2025-05-30 Diego Devesacuda : prevent using split buffers with 3d/4d matrices...
2025-05-30 Akarshan BiswasSYCL: Add mrope kernel (#13755)
2025-05-30 Georgi Gerganovsync : vendor (#13901)
2025-05-30 Sigbjørn Skjæretconvert : fix rwkv bos/eos token (#13844)
2025-05-30 Xuan-Son Nguyenconvert : allow partial update to the chkhsh pre-tokeni...
2025-05-30 Đinh Trọng Huyllama : add support for DistilBert (#13907)
2025-05-30 zhangkaihuollama : use llm_build_granite for minicpm (#13911)
2025-05-29 Christian Kastnercmake: Guard GGML_CPU_ALL_VARIANTS by architecture...
2025-05-29 Sigbjørn Skjæretllama : add support for jina-reranker-v2 (#13900)
2025-05-29 Sigbjørn Skjæretgguf-py : add support for sub_type (in arrays) in GGUFW...
2025-05-29 Yibo Caiarm64: optimize q4_k_q8_k kernel with i8mm (#13886)
2025-05-29 Christian Kastnercmake: Factor out CPU architecture detection (#13883)
2025-05-29 Vineel Abhinavggml: aarch64: Implement SVE F32 kernels for Mamba...
2025-05-29 Georgi Gerganovtests : remove json.hpp from a test (#13880)
2025-05-29 Sigbjørn Skjæretconvert : workaround for AutoConfig dummy labels (...
2025-05-29 Sigbjørn Skjæretllama : add RobertaForSequenceClassification reranker...
2025-05-29 Vineel Abhinavggml: aarch64: Implement SVE F32 kernels for vector...
2025-05-28 Beinseziigguf-py : fix SafetensorRemote return on undefined...
2025-05-28 Xuan-Son Nguyenllama : fix KV shift for qwen2vl (#13870)
2025-05-28 Xuan-Son Nguyenmtmd : move helpers to dedicated library (⚠️ breaking...
2025-05-28 bandotici: disable LLAMA_CURL for Linux cross-builds (#13871)
2025-05-28 Đinh Trọng Huyllama : add support for BertForSequenceClassification...
2025-05-28 Đinh Trọng Huyconvert: small addition to support LlamaModel (#13838)
2025-05-28 Skyserver: fix remove 'image_url'/'input_audio' json-objec...
2025-05-28 Xuan-Son Nguyenconvert : fix qwen omni conversion (#13859)
2025-05-28 Alex Fanthometests : change umlaut test (#11600)
2025-05-28 Johannes GäßlerCUDA: fix FA tg at long context for CC >= 8.9 (#13852)
2025-05-28 Xuan-Son Nguyenconvert : fix tensor naming conflict for llama 4 vision...
2025-05-28 leo-ponyCANN: Add SOC TYPE printing in cmake configuration...
2025-05-27 lhezopencl: add new ops - `argsort`, `div`, `sub`, `addrows...
2025-05-27 lhezopencl: mark `mul_mat` `f32f32` as supporting non-conti...
2025-05-27 Jeff Bolzvulkan: use timestamp queries for GGML_VULKAN_PERF...
2025-05-27 Georgi Gerganovcmake : add llama-cparams.cpp to build (#13832)
2025-05-27 Akarshan BiswasSYCL: add gelu_erf kernel (#13749)
2025-05-27 Georgi Gerganovsync : ggml
2025-05-27 Xuan-Son Nguyenggml : add ggml_repeat_4d (#13824)
2025-05-27 xctanggml : riscv: add xtheadvector support (#13720)
2025-05-27 Xuan-Son Nguyenmtmd : support Qwen 2.5 Omni (input audio+vision, no...
2025-05-27 bandotidocs: remove link for llama-cli function calling (...
2025-05-27 Christian Kastnerggml-cpu: x86 feature detection is specific to x86...
2025-05-27 Diego Devesaggml : allow CUDA graphs when using pipeline parallelis...
2025-05-27 Georgi Gerganovkv-cells : track min/max used cells and per-sequence...
2025-05-27 Georgi Gerganovsampling : make sure samplers return at least 1 token...
2025-05-27 Georgi Gerganovllama : validate seq id batch input (#13809)
2025-05-26 Olivier Chafikserver: --offline mode (#13804)
2025-05-26 Georgi Gerganovscripts : add option to compare commits in Debug (...
2025-05-26 Georgi Gerganovcuda : avoid cuGetErrorString (#13791)
2025-05-26 Akarshan BiswasSYCL: Add non contiguous support in RMS_NORM and NORM...
2025-05-26 Olivier Chafikserver: fix streaming crashes (#13786)
2025-05-26 standby24x7examples/training: Fix file name in README (#13803)
2025-05-26 Olivier Chafik`server`: fix format of streamed tool call deltas ...
2025-05-26 Olivier Chafikserver: fix regression on streamed non-chat completion...
2025-05-26 Georgi Gerganovexamples : allow extracting embeddings from decoder...
2025-05-26 Georgi Gerganovllama : clarify deprecation message (#13794)
2025-05-26 Romain Biessysycl: Add more debug prints (#13640)
2025-05-26 Jeff Bolzvulkan: mark IM2COL as supporting non-contig (#13783)
2025-05-26 Bizhao ShiCANN: Add the basic supports of Flash Attention kernel...
2025-05-25 Olivier Chafik`server`: add `--reasoning-budget 0` to disable thinkin...
2025-05-25 Xuan-Son Nguyenwebui : bump max upload file size to 500MB (#13779)
2025-05-25 Sigbjørn Skjærettests : improve UGM tokenizer test coverage (#13773)
2025-05-25 Georgi Gerganovkv-cache : rework kv_cell (#13706)
2025-05-25 Percy Piperrpc : Fix build on OpenBSD (#13541)
2025-05-25 Xuan-Son Nguyenmtmd : add support for Qwen2-Audio and SeaLLM-Audio...
2025-05-25 ddpasadocs : add Moondream2 pre-quantized link (#13745)
2025-05-25 Olivier Chafikserver: fix/test add_generation_prompt (#13770) upstream/latest
2025-05-25 Piotr Jasiukajtisllama : add support for Qwen3 MoE tied word embeddings...
2025-05-25 Akarshan BiswasSYCL: revert "sycl: simplify bin_bcast_kernel (#13383...
2025-05-25 Olivier Chafik`server`: streaming of tool calls and thoughts when...
2025-05-24 Diego Devesareleases : bundle llvm omp library in windows release...
2025-05-24 Diego Devesareleases : enable openmp in windows cpu backend build...
2025-05-24 Diego Devesaggml-cpu : set openmp wait time if not set (#13758)
2025-05-24 0cc4mMove GLM4 f32 attention fix to the correct function...
2025-05-24 Xuan-Son Nguyenggml : add ggml_gelu_erf() CUDA kernel (#13719)
2025-05-24 Sigbjørn Skjæretvocab : fix ugm tokenizer precision (#13743)
2025-05-24 Johannes GäßlerCUDA: fix race condition in FA vector kernels (#13742)
2025-05-23 Diego Devesaci : enable winget package updates (#13734)
2025-05-23 Diego Devesaci : add winget package updater (#13732)
2025-05-23 Georgi Gerganovhparams : initialize arrays (#13728)
2025-05-23 Xuan-Son Nguyenllama : allow custom list of swa_layers (#13726)
2025-05-23 Xuan-Son Nguyenserver : support audio input (#13714)
2025-05-23 Chenguang LiCANN: Support MUL_MAT_ID for q8_0 and q4_0 (#13705)
2025-05-23 Xuan-Son Nguyenggml : fix the order of ggml_unary_op (#13718)
2025-05-23 Jeff Bolzvulkan: support CPY from any type to itself (#13695)
2025-05-23 Jeff Bolzvulkan: Disable coopmat/coopmat2/bfloat extensions...
2025-05-23 Judduse LOG_WARN to replace `std::cerr` (#13657)
2025-05-22 Diego Devesarelease : fix windows hip release (#13707)
2025-05-22 Georgi Gerganovtts : fix n_ubatch + make WavTokenizer cache-less ...
2025-05-22 Xuan-Son Nguyenmtmd : add ultravox audio input (#13623)
2025-05-22 Aaron Teocommon: Include torch package for s390x (#13699)
2025-05-22 Georgi Gerganovserver : pad small embedding batches (#13692)
next