Skip to content

Get command/title embeddings locally #15

@forrestbao

Description

@forrestbao

We do not wanna package a sentence-BERT model into the SmartCommand plug-in. Nor do we wanna package the embeddings of commands into it too. Too many floating point numbers. They will make the plug-in too bulky. This is the last resort.

Three options left:

  1. Download Sentence-BERT model (or whatever Transformer-based model) and run a Python script to embed locally.
  2. Sending strings to a cloud server to embed. The cloud server can be ours or provided by a third party, such as OpenAI's Plugin API.
  3. Pre-embed the commands (and their labels/titles) and store in our cloud server. When SmartCommand is first installed, download such embeddings. We can expand the plug-ins supported gradually.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions