]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2025-04-11 Xuan-Son Nguyenclip : use smart pointer (⚠️ breaking change) (#12869)
2025-04-11 Akarshan BiswasSYCL: Add fp16 type support to unary op kernels (#12788)
2025-04-11 Daniel Hanconvert : Llama4 RoPE fix (#12889)
2025-04-11 R0CKSTARci : Replace freediskspace to free_disk_space in docker...
2025-04-11 Daniel Beveniusxcf : add check for visionos build version (#12854)
2025-04-11 Xuan-Son Nguyenconvert : proper tensor name mapping for llama4 (#12870)
2025-04-11 Xuan-Son Nguyenllama : correct rms norm for llama 4 (#12882)
2025-04-11 Aaron Teoggml: fix compilation error s390x (#12848)
2025-04-10 Georgi Gerganovsync : ggml
2025-04-10 Georgi Gerganovtests : fix init order (#0)
2025-04-10 Georgi Gerganovsync : ggml
2025-04-10 cmdr2ggml: don't include arm_neon.h when using CUDA 12 with...
2025-04-10 Diego Devesaggml : add bilinear upscale support (ggml/1185)
2025-04-10 Diego Devesaggml : add more generic custom op, remove deprecated...
2025-04-10 Georgi Gerganovscripts : fix sync-ggml-am.sh
2025-04-10 Xuan-Son Nguyenllava : introduce libmtmd (#12849)
2025-04-10 Xuan-Son Nguyenconvert : ability to lazy-load safetensors remotely...
2025-04-10 Chenguang LiCANN: Support more ops (#12841)
2025-04-09 Prajwal B MehendarkarFixes #12823 (#12830)
2025-04-09 Rudi Servodocker : added all CPU to GPU images (#12749)
2025-04-09 Piotr Kubajggml-cpu-impl.h: do not redefine bool on POWER9 (#12856)
2025-04-09 Piotr Kubajggml-impl.h: fix build on POWER9 (#12855)
2025-04-09 Bo Zhengllama : Support Qwen3 and Qwen3MoE (#12828)
2025-04-09 R0CKSTARmusa: enable freediskspace for docker image build ...
2025-04-09 Romain Biessysycl: update documentation to use -no-cnv (#12845)
2025-04-09 Plamen Minevci: detach common from the library (#12827)
2025-04-09 Xuan-Son Nguyenclip : do not print ftype (#12832)
2025-04-09 Georgi Gerganovreadme : add rpc backend (#12842)
2025-04-09 Chenguang LiCANN: Support Opt CONV_TRANSPOSE_1D and ELU (#12786)
2025-04-09 Jeff Bolzvulkan: In coopmat2 mmq, load q4_k/q5_k scales through...
2025-04-09 Jeff Bolzvulkan: Use fp16 for the flash attention P*V multiplica...
2025-04-08 Sigbjørn Skjæretcuda : add f32 to bf16 copy op (#12806)
2025-04-08 Matt Claytonllava: improve clip_ctx destructor to not memleak load_...
2025-04-08 Georgi Gerganovllama : fix FA when KV cache is not used (i.e. embeddin...
2025-04-08 Xuan-Son Nguyenserver : fix thread.join() on exit (#12831)
2025-04-08 dm4llava: add more helper functions to check projector...
2025-04-08 Prajwal B Mehendarkararg : Including limits file on AIX (#12822)
2025-04-08 characharmserver : webui : Improve Chat Input with Auto-Sizing...
2025-04-08 Neo Zhang JianyuRevert "sycl:remove redundant memcopy in function ggml_...
2025-04-08 compiladegguf-py : support lazy tensor splitting (#12809)
2025-04-07 Xuan-Son Nguyenllama : Support llama 4 text-only (#12791)
2025-04-07 lhezopencl: better identify Adreno GPU (#12760)
2025-04-07 stduhpfhellaswag: display estimated score confidence interval...
2025-04-07 Georgi Gerganovcuda : fix HIP and MUSA BF16 (#0)
2025-04-07 Georgi Gerganovsync : ggml
2025-04-07 Georgi Gerganovggml : simplify Arm fp16 CPU logic (ggml/1177)
2025-04-07 Sigbjørn SkjæretCUDA: don't convert BF16 weights to FP32 (ggml/1174)
2025-04-07 cmdr2cpu: move all the operators into a separate c++ file...
2025-04-07 zhouwgsycl: remove redundant memcopy in function ggml_backend...
2025-04-07 Xuan-Son Nguyenci : no curl on ggml-ci (#12796)
2025-04-07 Xuan-Son Nguyencmake : enable curl by default (#12761)
2025-04-07 zhouwgCANN: fix typo in ggml-cann (#12733)
2025-04-07 hipuddingCANN: Refactor to reduce duplicate code (#12731)
2025-04-06 R0CKSTARmusa: fix compilation warnings in mp_22/31 (#12780)
2025-04-06 Jeff Bolzvulkan: fix NaN issue in flash attention shader (#12776)
2025-04-06 Jeff Bolzvulkan: Use unclamped loads for flash attention mask...
2025-04-05 0cc4mVulkan: Tune Vulkan mmq int dot shader for performance...
2025-04-05 Sergey Fedorovcommon : fix includes in arg.cpp and gemma3-cli.cpp...
2025-04-05 Xuan-Son Nguyenclip : refactor clip_init, add tests (#12757)
2025-04-05 エシュナヴァリシアcommon: custom hf endpoint support (#12769)
2025-04-04 Olivier Chafiksync: minja (#12739)
2025-04-04 Georgi Gerganovkv-cache : simplify + fix warning for recurrent models...
2025-04-04 bandotici: add Linux cross-compile build (#12428)
2025-04-04 Nauful Shaikhserver : webui : Upgrade daisyui, tailwindcss. (#12735)
2025-04-04 nick huanggguf-split : --merge now respects --dry-run option...
2025-04-04 Nicolò Scipionesycl: allow ggml-sycl configuration and compilation...
2025-04-04 Ronny Brendelcmake: fix ggml-shaders-gen compiler paths containing...
2025-04-04 Daniel Beveniusdocs : add XCFramework section to README.md [no ci...
2025-04-04 Jeff Bolzvulkan: Hybrid waitForFences/getFenceStatus to reduce...
2025-04-04 Jeff Bolzvulkan: set cmake minimum and project name in vulkan...
2025-04-04 lhezopencl: update doc for OpenCL (#12702)
2025-04-03 Gaurav GargCUDA: Prefer vector flash decoding kernel for Gemma...
2025-04-03 yumeyaovocab : use string_view::find() to avoid unnecessary...
2025-04-03 Jeff Bolzvulkan: Fix missing cmake logic for dot product extensi...
2025-04-03 Atharva Dubeyci : add env variable in ggml-ci and document the same...
2025-04-03 R0CKSTARsync : minja (inclusionAI/Ling) and update tests (...
2025-04-03 a3shfix MUSA compiler warning (#12704)
2025-04-03 Chenguang LiCANN: Support operator SIN COS ARGMAX (#12709)
2025-04-03 Alan GraySimplify and improve CUDA graphs through use of indirec...
2025-04-03 hipuddingCANN: Fix failed test cases (#12708)
2025-04-03 lhezopencl: use `max_alloc_size` in backend ctx instead...
2025-04-02 Jeff Bolzvulkan: Implement split_k for coopmat2 flash attention...
2025-04-02 bandoticmake: remove caching from vulkan coopmat checks (...
2025-04-02 Jeff Bolzvulkan: Implement grouped query attention in the coopma...
2025-04-02 0cc4mVulkan: Fix mmq int dot float cache size (#12722)
2025-04-02 Georgi Gerganovmodel : print tensor size during load (#12711)
2025-04-02 Diego Devesallama : add option to override model tensor buffers... upstream/0.0.5028
2025-04-02 Georgi Gerganovllama : refactor kv cache guard (#12695)
2025-04-02 Sigbjørn Skjæretvocab : BailingMoE : change possessive quantifiers...
2025-04-02 Xuan-Son Nguyencommon : remove json.hpp from common.cpp (#12697)
2025-04-02 Chenguang Li[CANN] get_rows and dup optimization (#12671)
2025-04-01 Xuan-Son Nguyencommon : refactor downloading system, handle mmproj...
2025-04-01 Junil Kimopencl : fix memory allocation size (#12649)
2025-04-01 jklincnllama : use LLM_KV_GENERAL_FILE_TYPE instead of gguf_fi...
2025-04-01 Sigbjørn Skjæretconvert : BailingMoE : fix qkv split when head_dim...
2025-04-01 Georgi Gerganovmetal : use F32 prec in FA kernels (#12688)
2025-04-01 R0CKSTARFix clang warning in gguf_check_reserved_keys (#12686)
2025-04-01 Wagner Brunavulkan: fix build when glslc doesn't support coopmat...
2025-04-01 Romain BiessySYCL: Rename oneMKL to oneMath (#12192)
2025-04-01 Akarshan BiswasSYCL: switch to SYCL namespace (#12674)
next