tf-quant-finance

Leverages TensorFlow's GPU/TPU support for high-performance computations, as demonstrated in examples like Monte Carlo simulations and PDE solvers.

Organized into foundational, mid-level, and pricing methods with independent examples, making it accessible for various financial modeling tasks, as outlined in the introduction.

Utilizes TensorFlow's automatic differentiation for gradient-based methods, enabling efficient calibration and optimization, highlighted in the library's philosophy.

Provides end-to-end Jupyter notebooks for tasks like American option pricing and swap curve fitting, facilitating learning and practical application.