- A websocket based training/inference server.
- Flexible asyncio based pipeline execution framework.
- Make sure you have
redis-server
running at port 6379. If not, you can dodocker run -p 6379:6379 -d redis
. (Warning:test.py
run a flushall, it will erase your redis db!!!) - Setup environment by running
pipenv install --dev
- Enter the virtualenv by running
pipenv shell
- Inside of virtualenv, run:
python app.py
will start the trireme server.python test.py [interval]
will start sending requests. You can specify shorter interval in seconds for simulate real work load.
Checkout API.md