EvaDB provides couple different installation options to allow easy extension to rich functionalities.
By Default, EvaDB installs only the minimal requirements.
pip install evadb
You can install EvaDB with the vision extension. With vision extension, you can run queries to do image classification, object detection, and emotion analysis workloads, etc.
pip install evadb[vision]
Documents Summarization with LLM#
You can also use EvaDB to leverage the capability of LLM to summarize or do question answering for documents.
pip install evadb[document]
Additional Vector Index#
faiss vector index by default, but users can also install other index library such as
qdrant for similarity search feature.
pip install evadb[qdrant]
Training or Finetuning Model#
Instead of using existing models for only inference, you can also train a customized function inside EvaDB with the
pip install evadb[ludwig]
Better Performance and Scalability#
EvaDB also allows users to improve the query performance by using
ray to parallelize queries.
pip install evadb[ray]