This repository provides a collection of deep learning model implementations for time series forecasting built on TensorFlow and Keras. It includes architectures such as Transformers, WaveNet, and seq2seq that are adapted for sequential prediction tasks. The library enables developers to construct and train neural network models for forecasting future values from historical time series data.