This is a collection of Python libraries designed for fetching and formatting Pokémon-related information from the Smogon API, as well as implementing a neural network from scratch for embedding Pokémon data.
The goal of this project is to eventually use the pokemonembedding module to generate vectors for the strategy and store them in a vector database, which will be used with a LLM for strategy and information retrieval.
- API Module: Fetches and formats Pokémon-related information from the Smogon API.
- Embedding Module: Implements a neural network from scratch for embedding Pokémon data.
- Codec Implementation: Provides serialization and deserialization functionality for the neural network and its components.