Pinned Loading

1. BlockBlast-Game-AI-Agent BlockBlast-Game-AI-Agent Public

   BlockBlast reimplementation + RL agents (DQN, PPO, PPO+Action Masking, DQN+Action Masking, Random)

   Python 34 13

2. personalized-autocomplete-next-word-prediction-task personalized-autocomplete-next-word-prediction-task Public

   A personalized autocomplete (next word prediction) project using three different architectures: stacked LSTMs, Seq2Seq with Attention and LSTMs and GPT-2, written from scratch.

   Jupyter Notebook 10

3. ai-chess-opponent ai-chess-opponent Public

   AI chess opponent that plays reasonably well against humans (~1800-1900 ELO). The engine has 4 modes: random, minimax with piece valuation, minimax with custom piece evaluation heuristics + positio…

   Jupyter Notebook