Commits

Commits on Mar 19, 2023

Update llama_model_load() from master branch
thomasantony
committed
1c545e5
Copy full SHA for 1c545e5
Add support for multiple antiprompts
thomasantony
committed
5195fed
Copy full SHA for 5195fed
Update CMakeLists to rename targets
Show description for b3541ce
thomasantony
committed
b3541ce
Copy full SHA for b3541ce
Update main.cpp to use instruct mode
thomasantony
committed
1c1cf35
Copy full SHA for 1c1cf35
Update llama.cpp to use instruct mode
thomasantony
committed
edb52ab
Copy full SHA for edb52ab
Refactor interactive mode in main.cpp
thomasantony
committed
660a4d5
Copy full SHA for 660a4d5
Refactor llama.cpp and llama.h
thomasantony
committed
3839a08
Copy full SHA for 3839a08
Add llama_tokens_to_string() to utils.cpp
Show description for 05224ed
thomasantony
committed
05224ed
Copy full SHA for 05224ed
Apply suggestions to main.cpp
thomasantony
committed
912e624
Copy full SHA for 912e624
Apply suggestions to llama.cpp and llama.h
thomasantony
committed
b0ed03b
Copy full SHA for b0ed03b
Apply suggestions to CMakeLists.txt
thomasantony
committed
7fb4c51
Copy full SHA for 7fb4c51
Add make_unique for C++11
thomasantony
committed
a81b670
Copy full SHA for a81b670
Update main.cpp to use new llama library
thomasantony
committed
4b4d8a5
Copy full SHA for 4b4d8a5
Refactor code structure in llama.cpp and llama.h
thomasantony
committed
b14486e
Copy full SHA for b14486e
Add to CMakeLists.txt
thomasantony
committed
0995df5
Copy full SHA for 0995df5
Move llama_model_quantize() into llama.cpp
thomasantony
committed
ebfc622
Copy full SHA for ebfc622
Refactor out library code from main.cpp
thomasantony
committed
9c2109e
Copy full SHA for 9c2109e
Add llama.cpp and llama.h
thomasantony
committed
8f02f1b
Copy full SHA for 8f02f1b
Rename script
ggerganov
committed
4545539
Copy full SHA for 4545539
Add temporary helper script for Alpaca chat
ggerganov
committed
edeba28
Copy full SHA for edeba28
fix coloring of last `n_batch` of prompt, and refactor line input (#221)
Show description for 5c19c70
bitRAKE
authored
5c19c70
Copy full SHA for 5c19c70
Support for multiple reverse prompts. (#299)
Show description for 2456837

tjohnman
and
Johnman
authored
2456837
Copy full SHA for 2456837
Improved quantize script (#222)
Show description for 7392f1c
SuajCarrot
authored
7392f1c
Copy full SHA for 7392f1c
Make prompt randomization optional. (#300)
Show description for ad5fd5b
tjohnman
authored
ad5fd5b
Copy full SHA for ad5fd5b
Respect the maximum number of tokens in interactive. (#298)
Show description for 368d0c8

authored
368d0c8
Copy full SHA for 368d0c8
Add --ignore-eos parameter (#181)
Show description for 50fae10

slaren
and
ggerganov
authored
50fae10
Copy full SHA for 50fae10
interactive mode: print '\n' in sigint_handler, this flush stdout thus ensure color reset. (#283)
mqy
authored
084e2f0
Copy full SHA for 084e2f0
Command line switch to use F16 for memory_k and memory_v (refactor of #154) (#294)
Show description for 0b366e7

Green-Sky
and
ty-everett
authored
0b366e7
Copy full SHA for 0b366e7
Update hot topics to mention Alpaca support
ggerganov
authored
160bfb2
Copy full SHA for 160bfb2
Fix off-by-one bug (#115)
ggerganov
committed
c494ed5
Copy full SHA for c494ed5
Fix python stuff (#109)
ggerganov
committed
c1c7026
Copy full SHA for c1c7026
Refactoring `convert-pth-to-ggml.py`: more concise and readable (#109)
Show description for 467b149

qunash
and
ggerganov
authored
467b149
Copy full SHA for 467b149
Drop trailing new line from file prompts (#80)
ggerganov
committed
70f01cb
Copy full SHA for 70f01cb
Add instruction for using Alpaca (#240)
ggerganov
authored
a4e63b7
Copy full SHA for a4e63b7
Add "--instruct" argument for usage with Alpaca (#240)
Show description for 9e17072
ggerganov
committed
9e17072
Copy full SHA for 9e17072