tjohnman
|
305ba6f0e6
Don't force immediate interactive without `-i` (#354)
|
2 years ago |
Georgi Gerganov
|
f5a77a629b
Introduce C-style API (#370)
|
2 years ago |
Fabio R. Sluzala
|
353ec251a4
We could use std::unordered_map over std::map (#305)
|
2 years ago |
Gary Linscott
|
486ae645fd
Compute perplexity over prompt (#270)
|
2 years ago |
anzz1
|
975d2cebf9
cmdline option for custom amount of model parts (--n_parts N) (#348)
|
2 years ago |
Georgi Gerganov
|
8f644a0a85
Change default repeat_penalty to 1.0
|
2 years ago |
Georgi Gerganov
|
eb34620aec
Add tokenizer test + revert to C++11 (#355)
|
2 years ago |
Mack Straight
|
a791a68b61
move file magic/version to header, print expected version (#319)
|
2 years ago |
Mack Straight
|
074bea2eb1
sentencepiece bpe compatible tokenizer (#252)
|
2 years ago |
tjohnman
|
24568371ae
Support for multiple reverse prompts. (#299)
|
2 years ago |
tjohnman
|
ad5fd5b60c
Make prompt randomization optional. (#300)
|
2 years ago |
slaren
|
50fae10d03
Add --ignore-eos parameter (#181)
|
2 years ago |
Erik Scholz
|
0b366e7357
Command line switch to use F16 for memory_k and memory_v (refactor of #154) (#294)
|
2 years ago |
Georgi Gerganov
|
9e1707218a
Add "--instruct" argument for usage with Alpaca (#240)
|
2 years ago |
Georgi Gerganov
|
4f54609110
Default to 4 threads (#243)
|
2 years ago |
Stephan Walter
|
367946c668
Don't tell users to use a bad number of threads (#243)
|
2 years ago |
Justin Suess
|
2d64715ad4
added ctx_size parameter (#148)
|
2 years ago |
Matvey Soloviev
|
96ea727f47
Add interactive mode (#61)
|
2 years ago |
beiller
|
02f0c6fe7f
Add back top_k (#56)
|
2 years ago |
beiller
|
129c7d1ea8
Add repetition penalty (#20)
|
2 years ago |
Georgi Gerganov
|
70bc0b8b15
Fix a bug in the rope calculation
|
2 years ago |
Georgi Gerganov
|
319cdb3e1f
Final touches
|
2 years ago |
Georgi Gerganov
|
26c0846629
Initial release
|
2 years ago |