A web server for accessing generative language models from a smart speaker
To evaluate one message and generate response, use the ask
command:
python -m rt ask 'Есть такой анекдот' -c openai -m 'gpt-4'
To run http
server which would allow you to interact with the smart speaker and generate responses using external models, use the following command:
python -m rt serve -c openchat
To forward requests from localhost:2222
to foo.bar:3333
:
ssh -p 1111 -L 2222:localhost:3333 user@foo.bar
To run ngrok:
ngrok http --domain=foo-bar.ngrok-free.app 2222