Commit History

Autor SHA1 Mensaxe Data
  Georgi Gerganov c90d135eb4 examples : fix underscore in beam-search + .gitignore (close #2900) %!s(int64=2) %!d(string=hai) anos
  M. Yusuf Sarıgöz 0d1c706181 gguf : add workflow for Pypi publishing (#2896) %!s(int64=2) %!d(string=hai) anos
  alonfaraj 9509294420 make : add test and update CI (#2897) %!s(int64=2) %!d(string=hai) anos
  Gilad S 35092fb547 docs : add `node-llama-cpp` to `README.md` (#2885) %!s(int64=2) %!d(string=hai) anos
  Kerfuffle dc07dc492e convert : various script cleanups/fixes + merges and special token handling (#2842) %!s(int64=2) %!d(string=hai) anos
  chaihahaha ad9ddcff6e llm.vim : stop generation at multiple linebreaks, bind to <F2> (#2879) %!s(int64=2) %!d(string=hai) anos
  staviq 8341a25957 main : log file (#2748) %!s(int64=2) %!d(string=hai) anos
  Cebtenzzre 849408957c tests : add a C compliance test (#2848) %!s(int64=2) %!d(string=hai) anos
  slaren 06abf8eeba ggml : add view_src and view_offs to ggml_tensor for views (#2874) %!s(int64=2) %!d(string=hai) anos
  slaren c03a243abf remove outdated references to -eps and -gqa from README (#2881) %!s(int64=2) %!d(string=hai) anos
  Kawrakow fa3582f509 Tell users attmepting to run perplexity with too few tokens to use more (#2882) %!s(int64=2) %!d(string=hai) anos
  Kawrakow e37e69dcc3 10X faster BPE tokenizer (#2876) %!s(int64=2) %!d(string=hai) anos
  maddes8cht 53885d7256 py : fix "usage" messages (#2873) %!s(int64=2) %!d(string=hai) anos
  jameswu2014 bcce96ba4d convert.py : fix baichuan7B support (#2870) %!s(int64=2) %!d(string=hai) anos
  Jhen-Jie Hong 74e0caeb82 readme : add react-native binding (#2869) %!s(int64=2) %!d(string=hai) anos
  Cebtenzzre d4b5e16c32 make : fix clang tests build, add missing examples (#2859) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 3a007648f2 metal : add option to disable debug logs (close #2764) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 611363ac79 scripts : add pipefail %!s(int64=2) %!d(string=hai) anos
  Marcus Dunn 95b6e5212f added `struct` to llama_dump_timing_info_yaml's `llama_context` (#2857) %!s(int64=2) %!d(string=hai) anos
  xaedes 44c117f41e train : mem usage and other improvements (#2439) %!s(int64=2) %!d(string=hai) anos
  slaren 43033b7bb4 llama-bench : set locale to utf8 (#2832) %!s(int64=2) %!d(string=hai) anos
  Johannes Gäßler 6b73ef1201 YAML result logging + preset script (#2657) %!s(int64=2) %!d(string=hai) anos
  alonfaraj 75fafcbccc make : fix tests build (#2855) %!s(int64=2) %!d(string=hai) anos
  grahameth be475f60af llama.cpp : fix wrong vsnprintf call in MS compiler (#2856) %!s(int64=2) %!d(string=hai) anos
  Ronny Brendel 3af6b86301 ggml : tiny ggml_vec_dot_q4_K_q8_K AVX2 improvement (#2819) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov 35feac6560 ggml : sync (mem align to header + conv_transpose_2d fixes + ggml_alloc) (#2852) %!s(int64=2) %!d(string=hai) anos
  Johannes Gäßler 92b1bbd2ec CUDA: fix RoPE asserts, block sizes (#2833) %!s(int64=2) %!d(string=hai) anos
  igarnier dd0dc366da llama.h : add missing struct keyword for C compat in callback type (#2847) %!s(int64=2) %!d(string=hai) anos
  Georgi Gerganov f55538c3cc metal : fix memory leak (#2762) %!s(int64=2) %!d(string=hai) anos
  Cebtenzzre ebcee207b6 quantize : make output filename optional again (#2823) %!s(int64=2) %!d(string=hai) anos