cturan/llama.cpp

réplica de https://github.com/cturan/llama.cpp

Autor	SHA1 Mensaxe	Data
goerch	ff5a3f0c09 Work on the BPE tokenizer (#3252)	%!s(int64=2) %!d(string=hai) anos
cebtenzzre	1c84003c08 convert : fix vocab size when not defined in hparams (#3421)	%!s(int64=2) %!d(string=hai) anos
Cebtenzzre	ecf90b1a51 gguf : make token scores and types optional (#3347)	%!s(int64=2) %!d(string=hai) anos
Meng Zhang	4fe09dfe66 llama : add support for StarCoder model architectures (#3187)	%!s(int64=2) %!d(string=hai) anos