Doesn't appear to work correctly when loaded in a Docker container: if multiple instances of an image get started (e.g. as a swarm service), neither containers seem to be able to load the word2vec model.
There are other ways to get around this, but for now it's not critical - we can just load the 4 GB module for each worker independently.
Doesn't appear to work correctly when loaded in a Docker container: if multiple instances of an image get started (e.g. as a swarm service), neither containers seem to be able to load the word2vec model.
There are other ways to get around this, but for now it's not critical - we can just load the 4 GB module for each worker independently.
Ping @rahulbot.