Commit History

Autor SHA1 Mensaxe Data
  Ivan Komarov c12b14b77f benchmark : fix result validation in benchmark-q4_0-matmult (#987) %!s(int64=2) %!d(string=hai) anos
  katsu560 106faaf297 cmake : add finding the OpenBLAS header file (#992) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak c85e03d12e Revert "main : alternative instruct mode (Vicuna support, etc.) (#863)" (#982) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak 489093548c py : bump sentencepiece to 0.1.98 to support Python 3.11 (#976) %!s(int64=2) %!d(string=hai) anos
  Stephan Walter 93265e988a make : fix dependencies, use auto variables (#983) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak c56b715269 Expose type name from ggml (#970) %!s(int64=2) %!d(string=hai) anos
  Tomáš Pazdiora f4d277ae17 main : alternative instruct mode (Vicuna support, etc.) (#863) %!s(int64=2) %!d(string=hai) anos
  Kerfuffle c9a59b70a5 ggml : add unary and binary map operations (#874) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak a32f7acc9f py : cleanup dependencies (#962) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak 43ffdefb74 py : fix flake8 and isort nitpicks (#960) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 1623a6e9b4 ggml : minor %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov c14e0d2f23 ggml : always allocate buffers with size multiple of GGML_MEM_ALIGN %!s(int64=2) %!d(string=hai) anos
  comex 723dac55fa py : new conversion script (#545) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 0f07cacb05 ggml : fix q4_1 dot product types %!s(int64=2) %!d(string=hai) anos
  Howard Su c5d70f5c9e ggml : optimize rope function to avoid call powf in the tight loop (#807) %!s(int64=2) %!d(string=hai) anos
  Gary Linscott be87b6ed20 perplexity : add support for batch size to `--perplexity` (#407) %!s(int64=2) %!d(string=hai) anos
  CRD716 0e07e6a839 common : remove unnecessary includes (#947) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov a3a2a0eda8 ggml : add GGML_DEFAULT_N_THREADS %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov d990e3fffc ggml : speed-up ggml_vec_dot_q4_1() ARM_NEON + 32-bit ARM support (#900) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 9190e8eac8 llama : merge llama_internal.h into llama.h %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov c85980acd0 gitignore : benchmark %!s(int64=2) %!d(string=hai) anos
  Stephan Walter 6232f2d7fd ggml : optimize non-SIMD Q4_0 vector dot product (#703) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak 6c248707f5 ggml : introduce GGML_ALIGNED_MALLOC/GGML_ALIGNED_FREE macros (#884) %!s(int64=2) %!d(string=hai) anos
  CRD716 8cda5c981d fix whitespace (#944) %!s(int64=2) %!d(string=hai) anos
  CRD716 ec29272175 readme : remove python 3.10 warning (#929) %!s(int64=2) %!d(string=hai) anos
  Genkagaku.GPT 7e941b95eb readme : llama node binding (#911) %!s(int64=2) %!d(string=hai) anos
  Pavol Rusnak c729ff730a flake.nix: add all binaries from bin (#848) %!s(int64=2) %!d(string=hai) anos
  Judd 4579af95e8 zig : update build.zig (#872) %!s(int64=2) %!d(string=hai) anos
  Vladimir 8c3ffc2f04 ggml : update cblas_sgemm columns var to be more reasonable (#838) %!s(int64=2) %!d(string=hai) anos
  niansa/tuxifan 107980d970 examples : add -n to alpaca and gpt4all scripts (#706) %!s(int64=2) %!d(string=hai) anos