]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2025-11-27 Aclyvulkan : move contiguous checks to device_supports_op...
2025-11-27 Jeff Bolzvulkan: use a fixed 1KB buffer for the add_rms_fusion...
2025-11-27 Xuan-Son Nguyenserver: enable jinja by default, update docs (#17524)
2025-11-26 lhezopencl: add sqr, sqrt, mean and ssm_conv (#17476)
2025-11-26 Alberto Cabrera... Fix chunks being too small with small matrix sizes...
2025-11-26 Han Qingzheclip: (minicpmv) fix resampler kq_scale (#17516)
2025-11-26 Jeff Bolzvulkan: allow graph_optimize for prompt processing...
2025-11-26 Jeff Bolzvulkan: Implement top-k (#17418)
2025-11-26 xctanggml-cpu : add RISC-V Zvfh impl for ggml_vec_mad_f16...
2025-11-26 Adrien Gallouëtcmake : use EXCLUDE_FROM_ALL to avoid patch-boringssl...
2025-11-26 Adrien Gallouëtggml : fix ARM feature verification (#17519)
2025-11-26 Jiacheng (Jason... HIP: Patch failed testcase in WMMA-MMQ kernels for...
2025-11-26 hipuddingCANN: Add MROPE and IMROPE support (#17401)
2025-11-26 o7sichore: upgrade cpp-httplib from v0.27.0 to v0.28.0...
2025-11-26 Jeff Bolzvulkan: Implement GGML_OP_CUMSUM (#17479)
2025-11-25 Georgi Gerganovggml : add ggml_top_k (#17365)
2025-11-25 Aleksei Nikiforovconvert : fix big-endian conversion (#17431)
2025-11-25 Diego Devesacodeowners : remove slaren (#17492)
2025-11-25 TianHao324CANN: supports out_prod operator for F32 and F16 (...
2025-11-25 Pascalwebui: add rehype plugin to restore HTML in Markdown...
2025-11-25 Jeff Bolzvulkan: Use fewer rows for scalar FA when HS is not...
2025-11-25 Aaron Teollama: introduce support for model-embedded sampling...
2025-11-24 Jeff Bolzvulkan: more FA details in vk_perf_logger (#17443)
2025-11-24 Daniel Beveniusllama : skip output reordering for single token batches...
2025-11-24 Jiacheng (Jason... HIP: WMMA-MMQ kernels for RDNA 4 (#17156)
2025-11-24 Sigbjørn Skjæretconvert : allow quantizing lora again (#17453)
2025-11-24 Xuan-Son Nguyenserver: split server.cpp code into server/common/task...
2025-11-24 Daniel Beveniusexamples : add -kvu to batched usage example [no ci...
2025-11-24 Georgi Gerganovsync : ggml
2025-11-24 Daniel Beveniusggml : remove dirty flag from version string (ggml...
2025-11-24 Alberto Cabrera... ggml-cpu: arm64: q4_K repack gemm and gemv implementati...
2025-11-24 ixgbeggml: add RISC-V cpu-feats (#17461)
2025-11-24 william panmodels : Added support for RND1 Diffusion Language...
2025-11-24 Max Krasnyanskyhexagon: add support for ROPE_NEOX (#17458)
2025-11-24 Raul TorresCANN: Define `cann_graph_update_required` before macro...
2025-11-24 M. Mediouniggml-hexagon: Initial Hexagon v68/v69 support (#17394)
2025-11-23 nullnameggml-hexagon: add `hex_supported_buffer` for better...
2025-11-23 Pascalwebui: minor settings reorganization and add disable...
2025-11-23 Sigbjørn Skjæretcuda : support non-contiguous i32 to i32 copy (#17326)
2025-11-23 Eric Curtinvulkan: Update docker image to Ubuntu 26.04 to enable...
2025-11-23 Jeff Bolzvulkan: remove a couple unnecessary switches (#17419)
2025-11-22 Adrien Gallouëtci : switch to BoringSSL on Server workflow (#17441)
2025-11-22 Masato NakasakaRevive MUL_MAT_ID to perf testing (#17397)
2025-11-21 yuloHIP: RDNA4 tensor core support for MMF (#17077)
2025-11-21 lhezopencl: refine condition for kqv mm (#17392)
2025-11-21 ubergarmmodel : detect GigaChat3-10-A1.8B as deepseek lite...
2025-11-21 Adrien Gallouëtcmake : add option to build and link BoringSSL (#17205)
2025-11-21 Adrien Gallouëtci : start using OpenSSL (#17235)
2025-11-21 Jeff Bolzvulkan: disable async for older Intel devices (#17369)
2025-11-21 Raul TorresCANN: Refactor `evaluate_and_capture_cann_graph` (...
2025-11-20 nullnameggml-hexagon: fix swiglu failure at `test-backend-ops...
2025-11-20 Daniel Hanreadme : add Unsloth exporting to GGUF in tools (#17411)
2025-11-20 Xuan-Son Nguyengrammar: fix regression caused by #17381 (#17412)
2025-11-20 Aleksander... Improved file naming & structure for UI components...
2025-11-20 Piotr Wilkin... grammar : fix integer overflow (#17381)
2025-11-20 Georgi Gerganovsync : ggml
2025-11-20 YangLemetal : fix compile on macos 11 (whisper/3533)
2025-11-20 Georgi Gerganovcommon : more accurate sampling timing (#17382)
2025-11-20 o7siconvert : fix TypeError when loading base model remotel...
2025-11-20 Piotr Wilkin... ggml : Fix transposed SOLVE_TRI result (#17323)
2025-11-20 Scott FudallyDGX Spark: UMA support (#17368)
2025-11-20 Adrien Gallouëtggml : remove useless and error-prone variadic macros...
2025-11-20 sudhiarmkleidiai: fix zero-size array declaration (#17240)
2025-11-20 ixgbeggml-cpu:add RISC-V RVV (Zvfh) optimization for FP16...
2025-11-19 Giuseppe Scrivanovulkan: implement ADD1, ARANGE, FILL, SOFTPLUS, STEP...
2025-11-19 Jeff Bolzvulkan: support larger argsort (#17313)
2025-11-19 Jeff Bolzvulkan: Add copy_transpose shader (#17371)
2025-11-19 Aleksander... webui: Add a "Continue" Action for Assistant Message...
2025-11-19 Sigbjørn Skjæretconvert : use self.block_count everywhere instead of...
2025-11-19 Aman Guptacuda: fix rope fusion for gemma3 (#17378)
2025-11-19 Piotr Wilkin... Fix too relaxed check on CUDA "fast copy" (can_be_trans...
2025-11-19 Ruben Ortlamvulkan: force full subgroups for flash attention to...
2025-11-19 Jeremy Randggml-cpu: Don't pass -mpowerpc64 when -mcpu already...
2025-11-18 Xuan-Son Nguyenchat: fix int overflow, prevent size calculation in...
2025-11-18 Haiyue Wangvocab : call reserve() for building plamo-2-translate...
2025-11-18 hksdpc255common : Generalized XML-style tool-call parsing with...
2025-11-18 jiahao suci : change the openEuler-310p image to fix release...
2025-11-18 Georgi Gerganovgitignore : be more specific about ignored stuff (...
2025-11-18 Chenguang LiCANN: fix acl_tensor_ptr usage in ASCEND_310P ROPE...
2025-11-18 o7sifix: resolve undefined variable 'svr' compilation error...
2025-11-18 jiahao suCANN: Add openEuler-cann in build and release (#17192)
2025-11-18 Jeff Bolzvulkan: support noncontig i32 copy (#17328)
2025-11-17 Xuan-Son Nguyenserver: split HTTP into its own interface (#17216)
2025-11-17 Ruben Ortlamvulkan: add log RTE support to fix Nvidia CI (#17320)
2025-11-17 Adrien Gallouëtcmake : fix ARM feature verification (#17170)
2025-11-17 Adrien Gallouëtggml : add missing AVX512 feature checks (#17270)
2025-11-17 Georgi Gerganovmetal : support I32 -> I32 copy (#17317)
2025-11-17 Georgi Gerganovmetal : faster argsort (#17315)
2025-11-17 Georgi Gerganovmetal : add cumsum (#17305)
2025-11-17 hipuddingCANN: Use smart pointers to manage ACL objects (#17238)
2025-11-16 Pavels Zaicenkovsvulkan: add LOG operation support for F32 and F16 ...
2025-11-16 Ruben Ortlamvulkan: fix MMQ quantize_y condition (#17301)
2025-11-16 Eveci : revert #16249 (#17303)
2025-11-16 Georgi Gerganovmetal : remove obosolete asserts (#17295)
2025-11-16 Georgi Gerganovserver : handle context overflow during decode (#17267)
2025-11-16 lhezopencl: fix rms_norm_mul (#17250)
2025-11-16 shaofeiqiopencl: add kernel to handle mat mul in attention to...
2025-11-15 shani-fsycl : unify unary kernels with a generic implementatio...
2025-11-15 Aleksander... webui: Fix clickability around chat processing statisti...
2025-11-15 Pascalwebui: add OAI-Compat Harmony tool-call streaming visua...
next