Training & Evaluation Pipeline

[maximizemaxwell]

Issue 7 — Training & Evaluation Pipeline

Description: Implement training pipeline with loss, metrics, and evaluation.

What to do

• Implement BCE loss with logits
• Implement metrics: accuracy, F1
• Build training loop with Adam optimizer
• Add CLI arguments for model, epochs, lr, etc.
• Log training/validation results per epoch
• Evaluate on test set after training
• Save results to results/logs/

Where to work: training/simulator_runner.py, training/evaluation.py