How can I configure Apache Beam's ReadFromText to use windowing, as opposed to the default global windowing (in Python)?
I'm trying to ingest a 10gb file, and processing the large number of records in one big window is stalling my pipeline's throughput a lot.
