DQN Agent for Warehouse Storage Optimization (OpenAI Gym)

1. Project Overview:

   • This project implements a Deep Q-Network (DQN) to solve a simulated warehouse environment using PyTorch and TensorBoard for training visualization.
   • The objective of the algorithm is to optimize the usage of storage space in the simulated warehouse environment.

2. Features:

   • TensorBoard logging (reward mean, episode length mean, exploration rate, frames per second)
   • Model saving/loading for inference

3. Model Performance Graphs (TensorBoard):

   

   

4. Next Steps:

   • Setup Experience replay
   • Exploration Strategy to be confirmed
   • Create/Upload the following files:
     • requirements.txt
     • config.yaml
     • agent.py
     • train.py
     • evaluate.py
     • utils.py
     • dqn_model.pt
     • any other relevant .ipynb files
   • Upload TensorBoard logs