Commits

Commits on Jul 5, 2023

export llama_timings as struct and expose them in server
tobi
committed
efa86bf
Copy full SHA for efa86bf

Commits on Jul 4, 2023

basic response formatting
tobi
committed
c19daa4
Copy full SHA for c19daa4
fix mobile, fix missing prompt cache
tobi
committed
eee6d69
Copy full SHA for eee6d69
rework state management into session, expose historyTemplate to settings
tobi
committed
fedce00
Copy full SHA for fedce00
slightly nicer css
tobi
committed
98e612c
Copy full SHA for 98e612c
add /completion.js file to make it easy to use the server from js
tobi
committed
dd1df3f
Copy full SHA for dd1df3f
enable server in Makefiles
tobi
committed
8e1b04d
Copy full SHA for 8e1b04d
let's try this with the xxd tool instead and see if msvc is happier with that
tobi
committed
dc7dd08
Copy full SHA for dc7dd08
remove need for @microsoft/fetch-event-source dep (-7kb)
tobi
committed
34fc3c7
Copy full SHA for 34fc3c7
revert log format changes
tobi
committed
e192f95
Copy full SHA for e192f95
improvements
tobi
committed
0f95689
Copy full SHA for 0f95689
allow server to multithread
Show description for 7a38956
tobi
committed
7a38956
Copy full SHA for 7a38956
switched to fprintf logging and to access_log
tobi
committed
a30d4b2
Copy full SHA for a30d4b2
newline police
tobi
committed
c8cedf5
Copy full SHA for c8cedf5
embed index and add --path for choosing static dir
tobi
committed
022bf2b
Copy full SHA for 022bf2b
minor aesthetic fixes
tobi
committed
e3fba85
Copy full SHA for e3fba85
server : clear trailing whitespace

ggerganov
authored and
tobi
committed
c1cb0e1
Copy full SHA for c1cb0e1
tighter
tobi
committed
b07b271
Copy full SHA for b07b271
expose simple web interface on root domain
Show description for 627d3ba
tobi
committed
627d3ba
Copy full SHA for 627d3ba
Allow old Make to build server. (#2098)
Show description for acc111c
SlyEcho
authored
acc111c
Copy full SHA for acc111c
Update Makefile: clean simple (#2097)
ZhouYuChen
authored
23c7c6f
Copy full SHA for 23c7c6f

Commits on Jul 3, 2023

CI: make the brew update temporarily optional. (#2092)
Show description for 698efad
Green-Sky
authored
698efad
Copy full SHA for 698efad
[ggml] fix index for ne03 value in ggml_cl_mul_f32 (#2088)
gotope
authored
14a2cc7
Copy full SHA for 14a2cc7
fix server crashes (#2076)
SlyEcho
authored
1cf14cc
Copy full SHA for 1cf14cc
Fix crash of test-tokenizer-0 under Debug build (#2064)
Show description for cc45a7f
howard0su
authored
cc45a7f
Copy full SHA for cc45a7f
[llama] No need to check file version when loading vocab score (#2079)
howard0su
authored
55dbb91
Copy full SHA for 55dbb91

Commits on Jul 2, 2023

server: add option to output probabilities for completion (#1962)
Show description for d7d2e6a
WangHaoranRobin
authored
d7d2e6a
Copy full SHA for d7d2e6a
ggml : fix build with OpenBLAS (close #2066)
ggerganov
committed
46088f7
Copy full SHA for 46088f7

Commits on Jul 1, 2023

Better CUDA synchronization logic (#2057)
JohannesGaessler
authored
0bc2cdf
Copy full SHA for 0bc2cdf
Test-based VRAM scratch size + context adjustment (#2056)
JohannesGaessler
authored
befb3a3
Copy full SHA for befb3a3
cmake : don't force -mcpu=native on aarch64 (#2063)
Show description for b213227
dsd
authored
b213227
Copy full SHA for b213227
metal : release buffers when freeing metal context (#2062)
apage43
authored
2f8cd97
Copy full SHA for 2f8cd97
convert : add support of baichuan-7b (#2055)
Show description for 471aab6

foldl
and
Judd
authored
471aab6
Copy full SHA for 471aab6
llama : fix return value of llama_load_session_file_internal (#2022)
ggerganov
committed
463f2f4
Copy full SHA for 463f2f4
llama : catch llama_load_session_file_internal exceptions (#2022)
Show description for cb44dbc
randxie
authored
cb44dbc
Copy full SHA for cb44dbc