- multiple users
- role selection
- temperature selection
- gpt instance reset
- logging system
/start
Start conversation
/temp
Command opens menu to select temperature hyperparameter for gpt. Use low value for formal proofs, solving tasks and other technical questions. Use high value for more creative answers and non technical questions, like introductions and conclusions for articles and other places when you need more words. For more info you can ask gpt itself.
/role
Command asks to provede a new role in next message for gpt assistant. You can use roles like 'you are an article writing assistant' to help with article writing, or 'you are an interesting companion' for chatting, or even 'you are an pirate' to get pirate styled messages
/reset
Reset gpt instance and start conversation from begining
/help
Prints help message with brief description for all commands
python3
pip3
openai
python-telegram-bot
before first launch run install.sh
script to install needed modules for python:
sh install.sh
you need to place your openai token to file .openai.token
and telegram token to file .simplebot.token
in the root of repo
you can lauch bot simply by
python3 SimpleGPTbot.py
and it will write all logs to stdout
if you want to run bot as daemon - run daemon.sh
script:
sh daemon.sh
it will disconnect process from current shell (and you can close it) and write all logs to log.txt
file.
To stop daemon just find process with px aux
and kill
it:
kill $(ps aux | grep 'python3 SimpleGPTbot.py' | awk '{print $2}')