]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2026-04-21 Xuan-Son Nguyenmtmd: correct mtmd_decode_use_mrope() (#22188)
2026-04-21 Georgi Gerganovllama-ext : fix exports (#22202)
2026-04-21 Georgi Gerganovsync : ggml
2026-04-21 Georgi Gerganovggml : bump version to 0.10.0 (ggml/1463)
2026-04-21 Georgi Gerganovfit-params : refactor + add option to output estimated...
2026-04-21 xris99server : fix hardcoded proxy connection timeout in...
2026-04-20 leonardHONGggml-cuda: flush legacy pool on OOM and retry (#22155)
2026-04-20 Xuan-Son Nguyenmtmd: correct get_n_pos / get_decoder_pos (#22175)
2026-04-20 Georgi Gerganovserver : remove /api endpoints (#22165)
2026-04-20 Gaurav GargTensor-parallel: Fix delayed AllReduce on Gemma-4 MoE...
2026-04-20 Johannes GäßlerTP: fix 0-sized tensor slices, AllReduce fallback ...
2026-04-20 pl752ggml-cpu: Optimized x86 and generic cpu q1_0 dot (follo...
2026-04-20 neha-haggml-webgpu: updated matrix-vector multiplication ...
2026-04-20 Xuan-Son Nguyenmtmd: refactor mtmd_decode_use_mrope (#22161)
2026-04-20 SamareshSinghfix: GLM-DSA crash in llama-tokenize when using vocab_o...
2026-04-20 Georgi Gerganovserver : refactor "use checkpoint" logic (#22114)
2026-04-20 Katostrofik[SYCL] Fix reorder MMVQ assert on unaligned vocab sizes...
2026-04-20 Yes You Can... server: rename --clear-idle to --cache-idle-slots ...
2026-04-19 Alessandro... vendor : update cpp-httplib to 0.42.0 (#21781)
2026-04-19 Johannes GäßlerCUDA: refactor mma data loading for AMD (#22051)
2026-04-19 Aldehir Rojascommon/autoparser : allow space after tool call (#22073)
2026-04-19 uvosHIP: Remove unesscary NCCL_CHECK (#21914)
2026-04-19 Xuan-Son Nguyenmtmd: add pos_0 to mtmd_image_tokens_get_decoder_pos...
2026-04-19 Gaurav Gargggml : reduce CPU overhead in meta backend (#22041)
2026-04-19 Sigbjørn Skjæretci : install spirv-headers for vulkan-cross (#22109)
2026-04-19 Dowonconvert : support sentence-transformer 5.4 config files...
2026-04-19 texasichcmake: remove CMP0194 policy to restore MSVC builds...
2026-04-19 Sascha Rogmannserver : speculative checkpointing (#19493)
2026-04-19 Radoslav Gerganovrpc : refactor the RPC transport (#21998)
2026-04-18 Cetarthoriphrosserver: Expose `media_tag` on /props endpoint. (#22028)
2026-04-18 Sigbjørn Skjæretmodel : refactor bias tensor variable names (#22079)
2026-04-18 Sigbjørn Skjæretandroid : libcommon -> libllama-common (#22076)
2026-04-18 SamareshSinghggml-backend-meta: add multi-segment read support in...
2026-04-18 Sigbjørn Skjæretci : free disk space for rocm release (#22012)
2026-04-18 Sigbjørn Skjæretconvert : fix (ignore for now) typings errors (#22002)
2026-04-18 Johannes Gäßlerllama: fit ctx size for CPU only (#21568)
2026-04-17 Reese Levineggml-webgpu: fix compiler warnings and refactor FlashAt...
2026-04-17 Aman GuptaCUDA: use LRU based eviction for cuda graphs (#21611)
2026-04-17 Yuri Khrustalevci : add android arm64 build and release (#21647)
2026-04-17 65amtmd: add missing struct tag (#22023)
2026-04-17 Georgi Gerganovlibs : rename libcommon -> libllama-common (#21936)
2026-04-17 Eric Zhangmodel : Gemma4 model type detection (#22027)
2026-04-17 lhezopencl: refactor q8_0 set_tensor and mul_mat host side...
2026-04-16 Sigbjørn Skjæretcli : use get_media_marker (#22017)
2026-04-16 Xuan-Son Nguyencmake: use glob to collect src/models sources (#22005)
2026-04-16 nullnamehexagon: optimize HMX matmul operations (#21071)
2026-04-16 Xuan-Son Nguyenmodel: using single llm_build per arch (#21970)
2026-04-16 shaofeiqiopencl: add q5_K gemm and gemv kernels for Adreno ...
2026-04-16 Pascalserver: tests: fetch random media marker via /apply...
2026-04-16 PikaPikachumodel : refactor QKV into common build_qkv and create_t...
2026-04-16 Sigbjørn Skjæretmodel : support NVFP4 tensors for Gemma4 (#21971)
2026-04-16 Ruben Ortlamcodeowners: add team member comments (#21714)
2026-04-16 Anav PrasadConvert: Fix NemotronH Config Parsing (#21664)
2026-04-16 Aman Guptaggml: add graph_reused (#21764)
2026-04-16 Kusha Gharahimetal: Implement ROLL op (#21946)
2026-04-16 rehan-10xengineerggml-cpu: add 128-bit RVV implementation for Quantizati...
2026-04-16 rehan-10xengineerggml : implemented simd_gemm kernel for riscv vector...
2026-04-16 Yuannandevops : added spirv-headers to nix (#21965)
2026-04-16 Reese Levineggml-webgpu: compute pass batching and removing profili...
2026-04-16 Ludovic Henryci : Use ggml-org/ccache-action on RISC-V as well ...
2026-04-16 Katostrofik[SYCL] Fix Q8_0 reorder: garbage on 2nd prompt + crash...
2026-04-15 Xuan-Son Nguyenserver: use random media marker (#21962)
2026-04-15 Ruben Ortlamvulkan: optimize im2col (#21713)
2026-04-15 Pasha Khosravicuda: Q1_0 initial backend (#21629)
2026-04-15 Reese Levineggml-webgpu: Fix dequantization helpers to not pass...
2026-04-15 Johannes GäßlerCUDA: require explicit opt-in for P2P access (#21910)
2026-04-15 Johannes GäßlerCUDA: manage NCCL communicators in context (#21891)
2026-04-15 Valeriy Dubovrpc : add native RDMA transport for RPC backend (RoCEv2...
2026-04-15 Xuan-Son Nguyendocs: more extensive RoPE documentation [no ci] (#21953)
2026-04-15 Ruben Ortlamci: disable test-backend-ops on Vulkan llvmpipe run...
2026-04-15 Piotr Wilkin... autoparser: support case of JSON_NATIVE with per-call...
2026-04-15 Mattread n_ctx back after making llama_context (#21939)
2026-04-14 Yiwei Shaohexagon: optimization for HMX mat_mul (#21554)
2026-04-14 Xuan-Son Nguyenggml : remove ggml-ext.h (#21869)
2026-04-14 Georgi Gerganovmetal : fix FA support logic (#21898)
2026-04-14 Xuan-Son Nguyenmtmd: add mtmd_image_tokens_get_decoder_pos() API ...
2026-04-14 Jeff Bolzvulkan: Programmatically add RoundingModeRTE to all...
2026-04-14 Georgi Gerganovci : re-enable mac workflows (#21894)
2026-04-14 Seyoung Jeongmetal : add XIELU unary op (#20802)
2026-04-14 Adrien Gallouëtvendor : update BoringSSL to 0.20260413.0 (#21881)
2026-04-14 Richard Davisonggml : fix ARM NEON nvfp4 dot product on non-dotprod...
2026-04-14 texasichcmake: fix CMP0194 warning on Windows with MSVC (#21630)
2026-04-14 Reese Levineggml-webgpu: Update register tiling matmul to use f32...
2026-04-14 Berk Idemcommon: skip reasoning budget sampler when no budget...
2026-04-14 Jeff Bolzvulkan: Support GGML_TYPE_NVFP4 (#21455)
2026-04-14 Xuan-Son Nguyenserver: support OAI /v1/audio/transcriptions API (...
2026-04-13 Aldehir Rojascommon/gemma4 : handle parsing edge cases (#21760)
2026-04-13 Xuan-Son Nguyendocs: listing qwen3-asr and qwen3-omni as supported...
2026-04-13 Piotr Wilkin... chat: dedicated DeepSeek v3.2 parser + "official" templ...
2026-04-13 Christian Kastnerci: Also exempt 'security' tag from auto-close (#21844)
2026-04-13 Ruben Ortlamvulkan: Flash Attention DP4A shader for quantized KV...
2026-04-13 Adrien Gallouëtcommon : add download cancellation and temp file cleanu...
2026-04-13 Gaspard Petitserver: Expose build_info in router mode (#21835)
2026-04-13 Oliver SimonsCUDA: Limit DeviceSegmentedSort to immediate mode ...
2026-04-13 Xuan-Son Nguyenmtmd: use causal attn for gemma 4 audio (#21824)
2026-04-13 Rohan Jainwebui: add setting for first-line chat titles (#21797)
2026-04-13 Aleksander... webui: MCP Diagnostics improvements (#21803)
2026-04-13 Masashi YoshimuraRemove extra conditional check on debug mode. (#21798)
2026-04-13 Akarshan Biswassycl: disable Q1_0 in backend and cleanup unused variab...
2026-04-12 Sergiumtmd: fix crash when sending image under 2x2 pixels...
next