Are you running your own Echoprint server, that is, have you set up Solr and Tokyo Tyrant according to the instructions in the echoprint-server distribution (README.md)?
Once you have done so, ingesting the fingerprints into the server is a two-stage process. The overview is as follows:
- Use the echoprint-codegen to make fingerprints of your MP3s.
- Ingest the resulting fingerprints (JSON files) using the fastingest.py script included with echoprint-server.
The details are in the documentation referred to above, but please do let us know if any part of the process is unclear to you.