]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/shortlog
pkg/ggml/sources/llama.cpp
2024-01-31 Jared Van Bortelkompute : llama-bench support and ggml_cpu_has_kompute...
2024-01-30 Georgi GerganovRevert "server : change deps.sh xxd files to string...
2024-01-30 Georgi Gerganovserver : fix context shift (#5195)
2024-01-30 JohnnyBserver : change deps.sh xxd files to string literals...
2024-01-30 Kawrakowggml : fix IQ3_XXS on Metal (#5219)
2024-01-30 Georgi Gerganovsync : ggml (#0)
2024-01-30 Georgi Gerganovgguf : fix comparison (ggml/715)
2024-01-30 John Balis`ggml_cuda_cpy` support for 4d tensors and float16...
2024-01-30 Georgi Gerganovgguf : add input validation, prevent integer overflows...
2024-01-30 Georgi Gerganovci : fix yolo URLs + fix metal capture (ggml/712)
2024-01-30 Jack Mousseaumetal : add debug capture backend function (ggml/694)
2024-01-30 KawrakowFaster AVX2 dot product for IQ2_XS (#5187)
2024-01-30 KawrakowSOTA 3-bit quants (#5196)
2024-01-30 0cc4mVulkan Windows APU Memory Handling (#5199)
2024-01-30 Vladimir Malyutinquantize : fix typo (#5211)
2024-01-30 divinity76main : allow empty --prompt-cache file (#5176)
2024-01-30 Romain Neutronreadme : minor (#5204)
2024-01-30 Georgi Gerganovreadme : update hot topics
2024-01-30 Wu Jian Pingserver : improve README (#5209)
2024-01-29 Paul Tsochantarisggml alloc: Fix for null dereference on alloc failure...
2024-01-29 Jared Van Bortelkompute : fix fallback to CPU (#5201)
2024-01-29 Jared Van BortelNomic Vulkan backend (#4456)
2024-01-29 divinity76fix typo "RLIMIT_MLOCK" (#5175)
2024-01-29 Wu Jian Pingserver : embeddings compatibility for OpenAI (#5190)
2024-01-29 Georgi Gerganovpy : fix except (#5194)
2024-01-29 Sang-Kil Parkpy : improve BPE tokenizer support (#5189)
2024-01-29 slarenggml : add max buffer sizes to opencl and metal backend...
2024-01-29 Evecmake : fix Vulkan build (#5182)
2024-01-28 Paul Tsochantarismetal : free metal objects (#5161)
2024-01-28 Georgi Gerganovsync : ggml
2024-01-28 Georgi Gerganovggml : minor type fix (int64_t -> size_t)
2024-01-28 0cc4mggml : add Vulkan backend (#2059)
2024-01-28 Abhilash Majumderggml : add unified SYCL backend for Intel GPUs (#2690)
2024-01-28 Georgi Gerganovflake.lock: Update (#5162)
2024-01-28 Johannes GäßlerApply min_p to unsorted tokens (#5115)
2024-01-28 Johannes GäßlerTests for min_p, sampling queue (#5147)
2024-01-28 Marcus Dunnreadme : add link to rust bindings (#5148)
2024-01-28 sharpHLllama : add support for Orion-14B (#5118)
2024-01-28 Kyle Misteledocker : add server-first container images (#5157)
2024-01-27 Johnllava : support for Yi-VL and fix for mobileVLM (#5093)
2024-01-27 Georgi Gerganovsync : ggml
2024-01-27 Juddggml : check ggml_add src1 type (ggml/708)
2024-01-27 Michael KlimenkoRemove unused data and add fixes (#5154)
2024-01-27 Maximilian... server : add self-extend support (#5104)
2024-01-26 0cc4mAdd OpenCL add kernel (#5151)
2024-01-26 Jared Van Bortelcmake : pass CPU architecture flags to nvcc (#5146)
2024-01-26 slarencuda : fix tensor size calculation for non-split buffer...
2024-01-26 slarenggml-alloc : add 10% margin to the buffer sizes (#5149)
2024-01-26 snadampalggml : update softmax n_task calculation (#5126)
2024-01-26 Georgi Gerganovscripts : move run-with-preset.py from root to scripts...
2024-01-26 Georgi Gerganovtests : gitignore test-c.o
2024-01-26 Xuan Son Nguyenserver : refactored the task processing logic (#5065)
2024-01-26 crasmci : add model tests + script wrapper (#4586)
2024-01-26 Paul Tsochantarismetal : remove unused `n_buffers` and `buffers` (#5129)
2024-01-26 Riceball LEEgguf : fix "general.alignment" type in gguf_reader...
2024-01-26 Georgi Gerganovreadme : update hot topics
2024-01-26 KawrakowAnother bucket sort (#5109)
2024-01-25 XiaotaoChenreadme : add MobileVLM 1.7B/3B to the supported models...
2024-01-25 l3utterflyllama : dynamic temperature sampling (#4972)
2024-01-25 Jared Van Bortelexamples : make pydantic scripts pass mypy and support...
2024-01-25 Valentin Konovalovandroid : use release cmake build type by default ...
2024-01-25 KawrakowFix Q3_K_XS for MoE models (#5113)
2024-01-25 Georgi Gerganovmetal : show compile log messages
2024-01-24 Engininja2cuda : fix 2-bit quants on amd hip (#5105)
2024-01-24 Michael Hueschennix-shell: use addToSearchPath
2024-01-24 Michael Hueschennix: add cc to devShell LD_LIBRARY_PATH
2024-01-24 slarenllama : pre-allocate input tensors in a separate buffer...
2024-01-23 Georgi Gerganovmetal : disable support for MUL_MAT F32 x F16
2024-01-23 KawrakowAdditional KL-divergence statistics (#5081)
2024-01-23 Johannes GäßlerCUDA: more info when no device code (#5088)
2024-01-23 Georgi Gerganovminor : clean-up some warnings and style (#5094)
2024-01-23 Xuan Son Nguyendevops : add intel oneapi dockerfile (#5068)
2024-01-23 Michael Coppolallama.vim : added api key support (#5090)
2024-01-22 slarenllama : fix not enough space in buffer with Qwen (...
2024-01-22 KawrakowKL-divergence (#5076)
2024-01-22 Reinforce-IIggml : parallelize FP32 conversion when using BLAS...
2024-01-22 XiaotaoChenllava : MobileVLM support (#4954)
2024-01-22 Someone Sergeflake.nix: add a comment about flakes vs nix
2024-01-22 Someone Sergenix: add a comment on the many nixpkgs-with-cuda instances
2024-01-22 Someone Sergenix: add a comment about makeScope
2024-01-22 Someone Sergenix: refactor the cleanSource rules
2024-01-22 Someone Sergeworkflows: nix-ci: drop the redundant "paths" filter
2024-01-22 Someone Sergeworkflows: nix-build-aarch64: rate limit
2024-01-22 Someone Sergeworkflows: nix-ci: rebuild on flake.lock updates
2024-01-22 Kawrakowimatrix : keep intermediate imatrix results (#5077)
2024-01-22 compiladellama : support StableLM 2 1.6B (#5052)
2024-01-22 Daniel Beveniusfinetune : print sample-start/include-sample-start...
2024-01-22 Kawrakowllama : add Q3_K_XS (#5060)
2024-01-22 bobqianicci : fix Windows CI by updating Intel SDE version ...
2024-01-22 Shijiellama : add more qwen2 models (#5071)
2024-01-21 iSmaRevert LLAMA_NATIVE to OFF in flake.nix (#5066)
2024-01-21 kuronekosaikoadd safetensors support to convert-lora-to-ggml.py...
2024-01-21 bobqianicadd `#include <string>` to unicode.h (#5051)
2024-01-21 KawrakowAdd ability to evauate multiple choice tasks (#5047)
2024-01-21 KawrakowSlightly faster imatrix (#5050)
2024-01-21 Georgi Gerganovflake.lock: Update (#5054)
2024-01-20 Jared Van Bortelconvert : partially revert PR #4818 (#5041)
2024-01-20 Jared Van Bortelperplexity : fix MSVC build after #5020 (#5043)
2024-01-20 slarenllama : run all KQV ops on the CPU with no KV offload...
2024-01-20 Herman Semenovcmake : add support for ccache (#5002)
next