remove prompt caching

bump

map model name to template, map template to prefix/suffix

clear cache script

new assistant script is just a second profile

minor text changes

use llama.cpp server instead of llama-cpp-python

integrate Chat-UI

koboldcpp launch script

llama-server.sh logs to file

update for dolphin

changed name of llava-cli

scarlett support

no trailing space on assistant prompt

fix cache path

refactor

include speculative scripts but do not install

include training script but do not install

new llama-prompt script

update to installed paths

fix path for prompts

add some old prompts ... to get started

llava image describe

openai-compatible server

empty arg

shebang

llama general-runner

finetuning script

llama.cpp updater script

bugfixes

typo

expand home path

fix license, language

model card metadata

add script that converts HF torch models to gguf

initial commit

initial commit