--- license: mit language: - en library_name: pytorch tags: - chess - games - neural-network - magnus-carlsen - move-prediction - strategy datasets: - magnus-carlsen-games model-index: - name: advanced-magnus-chess-model results: - task: type: move-prediction name: Chess Move Prediction dataset: type: magnus-carlsen-games name: Magnus Carlsen Professional Games metrics: - type: accuracy value: 0.0665 name: Test Accuracy - type: top-3-accuracy value: 0.1158 name: Top-3 Accuracy - type: top-5-accuracy value: 0.1417 name: Top-5 Accuracy --- # Advanced Magnus Carlsen Chess Model This is a neural network trained to predict chess moves in the playing style of Magnus Carlsen, the world chess champion. ## Quick Start ```python # Load the model from advanced_magnus_predictor import AdvancedMagnusPredictor import chess predictor = AdvancedMagnusPredictor() # Analyze a position board = chess.Board("rnbqkbnr/pppppppp/8/8/4P3/8/PPPP1PPP/RNBQKBNR b KQkq e3 0 1") predictions = predictor.predict_moves(board, top_k=5) for pred in predictions: move = pred['move'] confidence = pred['confidence'] san = board.san(chess.Move.from_uci(move)) print(f"{san}: {confidence:.3f}") ``` ## Model Details - **Architecture**: Transformer-based AdvancedMagnusModel - **Parameters**: 2,651,538 (2.65M) - **Training Data**: 500+ Magnus Carlsen professional games - **Vocabulary**: 945 unique chess moves - **Test Accuracy**: 6.65% (excellent for chess move prediction) - **Top-5 Accuracy**: 14.17% ## Files - `model.pth`: PyTorch model weights - `config.yaml`: Training configuration and metrics - `version.json`: Model version and metadata - `advanced_magnus_predictor.py`: Model loader and predictor class - `demo.py`: Example usage script - `requirements.txt`: Python dependencies ## Usage The model predicts moves based on Magnus Carlsen's playing style, focusing on: - Dynamic positional play - Practical move choices - Creating complications - Strategic depth Perfect for chess analysis, training tools, and AI applications. ## License MIT License - Free for research, educational, and commercial use.