Add support for the [vectors](https://neuml.github.io/txtai/embeddings/configuration/vectors/#vectors_1) parameter, which allows passing model-specific parameters. This allows customizing the new multiprocessing parameters.