 l3utterfly
|
8d608a81b7
main : fix off by one error for context shift (#6921)
|
1 год назад |
 Daniel Bevenius
|
5539e6fdd1
main : fix typo in comment in main.cpp (#6985)
|
1 год назад |
 Johannes Gäßler
|
28103f4832
Server: fix seed for multiple slots (#6835)
|
1 год назад |
 Pedro Cuenca
|
b97bc3966e
llama : support Llama 3 HF conversion (#6745)
|
1 год назад |
 Jared Van Bortel
|
1b67731e18
BERT tokenizer fixes (#6498)
|
1 год назад |
 Jan Boon
|
beea6e1b16
llama : save and restore kv cache for single seq id (#6341)
|
1 год назад |
 Georgi Gerganov
|
05b06210c9
llama : more consistent names of count variables (#5994)
|
1 год назад |
 DAN™
|
5a51cc1bb4
main : support special tokens as reverse/anti prompt (#5847)
|
1 год назад |
|
Georgi Gerganov
|
bf08e00643
llama : refactor k-shift implementation + KV defragmentation (#5691)
|
1 год назад |
|
Jared Van Bortel
|
89febfed93
examples : do not assume BOS when shifting context (#5622)
|
1 год назад |
 bmwl
|
f486f6e1e5
ggml : add numa options (#5377)
|
1 год назад |
|
Georgi Gerganov
|
85910c5b30
main : ctrl+C print timing in non-interactive mode (#3873)
|
1 год назад |
 Michael Klimenko
|
52bb63c708
refactor : switch to emplace_back to avoid extra object (#5291)
|
1 год назад |
 divinity76
|
813416991a
main : allow empty --prompt-cache file (#5176)
|
1 год назад |
 Yann Follet
|
722d33f34e
main : add parameter --no-display-prompt (#4541)
|
2 лет назад |
|
Georgi Gerganov
|
7edefbd79c
main : better name for variable n_print (#4874)
|
2 лет назад |
|
Georgi Gerganov
|
3ca63b4538
main : disable token count by default (#4874)
|
2 лет назад |
 pudepiedj
|
43f76bf1c3
main : print total token count and tokens consumed so far (#4874)
|
2 лет назад |
|
Georgi Gerganov
|
52531fdff8
main : add self-extend support (#4815)
|
2 лет назад |
 MaggotHATE
|
52c8bc3cf3
sampling : custom samplers order (#4285)
|
2 лет назад |
 Andrew Godfrey
|
8efa0f6ebe
main : pass LOG_TEE callback to llama.cpp log (#4033)
|
2 лет назад |
 Seb C
|
881800d1f0
main : Add ChatML functionality to main example (#4046)
|
2 лет назад |
 Kerfuffle
|
91f6499393
Respect tokenizer.ggml.add_bos_token value when tokenizing (#4040)
|
2 лет назад |
 cebtenzzre
|
b12fa0d1c1
build : link against build info instead of compiling against it (#3879)
|
2 лет назад |
|
Kerfuffle
|
6e08281e58
Extend llama_kv_cache_seq_rm to allow matching any sequence (#3843)
|
2 лет назад |
 Marcus Dunn
|
5be6c803fa
llama : remove token functions with `context` args in favor of `model` (#3720)
|
2 лет назад |
 vvhg1
|
d3956aea53
main : escape prompt for cfg_negative_prompt and consecutive inputs in main with interactive (#3623)
|
2 лет назад |
|
Georgi Gerganov
|
d1031cf49c
sampling : refactor init to use llama_sampling_params (#3696)
|
2 лет назад |
|
Georgi Gerganov
|
0e89203b51
speculative : add tree-based sampling example (#3624)
|
2 лет назад |
|
Georgi Gerganov
|
e1675d133c
llama : avoid fprintf in favor of LLAMA_LOG (#3538)
|
2 лет назад |
