Vector Embedding

How I Made AI Assistants Do My Work For Me: CrewAIYouTube

What are Vector Embeddings

It is a popular technique to represent 'information' in a format that can be easily processed by algorithms, especially deep learning models. This 'information' can be text, pictures, video and audio.

Each word gets its own vector with a lot of numbers. And the vector depends on the Model that generated them.

• OpenAI

• Word2vec

• GloVe

These vectors help finding data similar to them.

What they are used for?

• Recommendation systems

• Anomaly detection

• Transfer learning

• Visualizations

• Information retrieval

• Audio + Speech processing

• Facial recognition

Langchain

An open source framework that allows AI developers to have better interactions with several large language models LLMs.

pip install cassio datasets langchain openai tiktoken

• cassio is a Database solution