In the last few days we've noticed intermittent read timeouts for ops.epo.org

HTTPSConnectionPool(host='ops.epo.org', port=443): Read timed out. (read timeout=10.0)

When looking into where the timeout value is coming from, I saw that there's a constant NETWORK_TIMEOUT defined in api.py which can not be changed easily.

I will prepare a PR that will make this setting configurable when instantiating the client, defaulting to the 10 second timeout as before.