 slaren
|
f4d973cecb
convert.py : fix llama/llama2 conversion due to vocab_size=-1 (#4258)
|
%!s(int64=2) %!d(string=hai) anos |
 crasm
|
3014b5415d
Update docs for yarn_ext_factor <0.0 as unspecified instead of NaN (#4189)
|
%!s(int64=2) %!d(string=hai) anos |
 Galunid
|
f23c0359a3
ci : add flake8 to github actions (python linting) (#4129)
|
%!s(int64=2) %!d(string=hai) anos |
 Don Mahurin
|
2ab0707acb
convert : use 'model' value if it exists. This allows karpathy/tinyllamas to load (#4089)
|
%!s(int64=2) %!d(string=hai) anos |
 afrideva
|
b46d12f86d
convert.py: also look for plain model.safetensors (#4043)
|
%!s(int64=2) %!d(string=hai) anos |
 Kerfuffle
|
34b0a08207
gguf-py: Refactor and allow reading/modifying existing GGUF files (#3981)
|
%!s(int64=2) %!d(string=hai) anos |
|
Galunid
|
a75fa576ab
scripts: Generalize convert scripts (#3838)
|
%!s(int64=2) %!d(string=hai) anos |
 cebtenzzre
|
898aeca90a
llama : implement YaRN RoPE scaling (#2268)
|
%!s(int64=2) %!d(string=hai) anos |
 Georgi Gerganov
|
8a2f2fea29
convert : ignore tokens if their IDs are within [0, vocab_size) (#3831)
|
%!s(int64=2) %!d(string=hai) anos |
|
Kerfuffle
|
a5e7dbd614
llama : validate special token ids are in range when loading GGUF model (#3635)
|
%!s(int64=2) %!d(string=hai) anos |
 Qin Yue Chen
|
8cf19d60dc
gguf : support big endian platform (#3552)
|
%!s(int64=2) %!d(string=hai) anos |
 goerch
|
ff5a3f0c09
Work on the BPE tokenizer (#3252)
|
%!s(int64=2) %!d(string=hai) anos |
|
cebtenzzre
|
0fe321031a
gguf : general usability improvements (#3409)
|
%!s(int64=2) %!d(string=hai) anos |
 Zhang Peiyuan
|
e519621010
convert : remove bug in convert.py permute function (#3364)
|
%!s(int64=2) %!d(string=hai) anos |
 Erik Scholz
|
6eeb4d9083
convert: remove most of the n_mult usage in convert.py (#3098)
|
%!s(int64=2) %!d(string=hai) anos |
|
Cebtenzzre
|
6336d834ec
convert : fix F32 ftype not being saved (#3048)
|
%!s(int64=2) %!d(string=hai) anos |
|
Erik Scholz
|
c9c3220c48
convert: fix convert.py not working with int filename_stem (#3028)
|
%!s(int64=2) %!d(string=hai) anos |
|
Kerfuffle
|
cff7b0bf07
convert.py : BPE fixes (#2938)
|
%!s(int64=2) %!d(string=hai) anos |
|
Cebtenzzre
|
bce1fef328
convert : fix another python 3.8 issue (#2949)
|
%!s(int64=2) %!d(string=hai) anos |
|
Kerfuffle
|
aeefac4ff7
scripts: Use local gguf package when running from repo (#2927)
|
%!s(int64=2) %!d(string=hai) anos |
|
Cebtenzzre
|
92d0b751a7
convert : fix python 3.8 support, modernize type annotations (#2916)
|
%!s(int64=2) %!d(string=hai) anos |
|
Georgi Gerganov
|
b532a69b2f
convert.py : use dir name to name the llama
|
%!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 |
 jameswu2014
|
bcce96ba4d
convert.py : fix baichuan7B support (#2870)
|
%!s(int64=2) %!d(string=hai) anos |
|
Kerfuffle
|
730d9c681e
convert.py : advanced option (#2753)
|
%!s(int64=2) %!d(string=hai) anos |
 Nigel Bosch
|
a2ca4e9de9
Handle null rope scaling value (#2793)
|
%!s(int64=2) %!d(string=hai) anos |
|
Nigel Bosch
|
28b2c996ca
convert.py : Get rope scale from HuggingFace models (#2772)
|
%!s(int64=2) %!d(string=hai) anos |
|
slaren
|
12e2e33a97
convert.py : export rope freq_base when converting CodeLlama from an HF model (#2773)
|
%!s(int64=2) %!d(string=hai) anos |
|
slaren
|
d0f77b1353
convert.py : try to determine n_ctx automatically for CodeLlama (#2770)
|
%!s(int64=2) %!d(string=hai) anos |
|
slaren
|
0d3094f0c7
gguf : add rope_freq_base parameter for CodeLlama (#2769)
|
%!s(int64=2) %!d(string=hai) anos |
