Where do models get stored on the file system?

[LondonAppDev]

I'm running OpenSearch using Docker Compose.

In order to persist the data, I map /usr/share/opensearch/data to a volume on the host. This works for storing indexed items in OpenSearch, however, it doesn't seem to cache the models loaded into the ml plugin.

Is there a separate location on the filesystem outside of /usr/share/opensearch/data where the plugin/model data is stored?

[LondonAppDev]

I figured out the issue.

Basically, every time I restart OpenSearch, I need to load the model gain. I expect this is because the model needs to be "loaded" into memory in order to use it.