ggerganov/llama.cpp

Pull Requests Commits

Gate signal support on being on a unixoid system. (#74)

blackhole89 committed 2 years ago

a169bb88

Fix token count accounting

blackhole89 committed 2 years ago

460c4825

Revert "10% performance boost on ARM"

ggerganov committed 2 years ago

Verified c80e2a8f

Check for vdotq_s32 availability

ggerganov committed 2 years ago

Verified 54a0e66e

Ammend to previous commit - forgot to update non-QRDMX branch

ggerganov committed 2 years ago

Verified 543c57e9

10% performance boost on ARM

ggerganov committed 2 years ago

Verified 113a9e83

Fix color getting reset before prompt output done (#65)

blackhole89 committed 2 years ago

Verified 404fac0d

Update README.md

ggerganov committed 2 years ago

Verified 1a0a7430

Add interactive mode (#61)

blackhole89 committed 2 years ago

Verified 96ea727f

Fix typo in README (#45)

marckohlbrugge committed 2 years ago

Verified 96619548

Allow using prompt files (#59)

bengarney committed 2 years ago

Verified f385f8de

Add back top_k (#56)

beiller committed 2 years ago

Verified 02f0c6fe

Windows fixes (#31)

etra0 committed 2 years ago

Verified eb062bb0

Update README.md

ggerganov committed 2 years ago

Verified 7027a978

ggerganov committed 2 years ago

Verified 2d555e5b

Revert "weights_only" arg - this causing more trouble than help

ggerganov committed 2 years ago

Verified 7c9e54e5

python/pytorch compat notes (#44)

wizzard0 committed 2 years ago

Verified b9bd1d01

Add repetition penalty (#20)

beiller committed 2 years ago

Verified 129c7d1e

Clarify meaning of hacking

ggerganov committed 2 years ago

Verified 702fddf5

README: add "Supported platforms" + update hot topics

ggerganov committed 2 years ago

Verified 7d86e25b

use weights_only in conversion script (#32)

deepdiffuser committed 2 years ago

Verified a9312023

Add LICENSE (#21)

prusnak committed 2 years ago

Verified 6a9a67f0

Update README.md

ggerganov committed 2 years ago

Verified da1a4ff0

Fix a typo in model name (#16)

jooray committed 2 years ago

Verified 6b2cb630

Update README.md

ggerganov committed 2 years ago

Verified 4235e3d5

Add AVX2 support for x86 architectures thanks to @Const-me !

ggerganov committed 2 years ago

f1eaff47

Fix un-initialized FP16 tables on x86 (#15, #2)

ggerganov committed 2 years ago

a9e58529

Bump memory buffer

ggerganov committed 2 years ago

Verified 7d9ed7b2

Update README.md

ggerganov committed 2 years ago

Verified 0c680332

.gitignore models/

ggerganov committed 2 years ago

Verified f60fa9e5

Older