cd 'C:\Users\kbdsj\Virtual Environments\fyp-rl'

.\Scripts\activate

cd C:\Users\kbdsj\Desktop\Projects\FYP\RL-model\dev

pip freeze > requirements.txt

Each episode is a complete training session where:

1. The agent starts with a random/QWERTY layout

2. Takes multiple actions (swaps) while learning

3. Updates its policy based on rewards