Optimising the cartpole problem using RL reinforcement learning. Notebook only works in Google Colab environment.
- gym: Provides access to various reinforcement learning environments.
- numpy: Used for numerical computations (arrays, matrices, etc.).
- collections: Provides data structures like deque used for storing experiences.
- random: Used for generating random numbers.
- matplotlib: Used for creating plots and visualizations.
- matplotlib.pyplot: Shortcut for commonly used functions from matplotlib.
- torch: Deep learning library (used for defining the RL agent's neural network).
- torch.nn: Submodule of torch providing building blocks for neural networks.
- torch.nn.functional: Submodule of torch providing functional implementations of common layers.
- torch.optim: Submodule of torch providing optimization algorithms for training the network.
- torch.distributions: Submodule of torch providing probability distributions (used for action selection).
- glob: Used for finding all files matching a specific pattern (used in show_video).
- io: Used for file operations (used in show_video).
- base64: Used for encoding and decoding binary data (used in show_video).
- IPython.display: Used for displaying content in Jupyter notebooks (used in show_video).