NLP Example flow based on mlrun/storey
The demo shows:
- Using storey classes
- Using MLRun's
flow
serving - Defining and Deploying a
Graph
over 2 functions or more functions
Please run nlp-pipe.ipynb
Stream Watcher utility was added.
To use please:
python -m stream_watcher --stream <container> <stream_path> <shard_id> <seek_to (EARLIEST/LATEST)> <polling_interval in seconds>
You can add multiple streams by specifying more --stream <container> <stream_path> <shard_id> <seek_to (EARLIEST/LATEST)>
.
V3IO_WEBAPI
, V3IO_ACCESS_KEY
, V3IO_USERNAME
will be taken from the available environment variables.