KraTUZen/q-Taxi-v3

🚖 Q-Learning Agent on Taxi-v3

This repository contains a trained Q-Learning agent that successfully solves the Taxi-v3 environment.

---

📊 Model Card

Model Name: q-Taxi-v3
Environment: Taxi-v3
Algorithm: Q-Learning
Performance Metric:

• Mean Reward: Achieves optimal policy performance in evaluation runs
• Verification: Not yet independently verified

---

🚀 Usage

from huggingface_hub import load_from_hub
import gym

# Load the trained Q-learning model
model = load_from_hub(
    repo_id="KraTUZen/q-Taxi-v3",
    filename="q-learning.pkl"
)

# Initialize environment
env = gym.make(model["env_id"])

---

🧠 Notes

• The agent is trained on the Taxi-v3 environment, where the goal is to pick up and drop off passengers efficiently.
• The Q-table is serialized in q-learning.pkl.
• You can directly load and evaluate the agent using the provided snippet.

---

📂 Repository Structure

• q-learning.pkl → Serialized Q-table of the trained agent
• README.md → Documentation and usage guide

---

✅ Results

The agent consistently learns an optimal policy for navigating the grid, picking up passengers, and dropping them at their destinations with minimal steps.

---

🔎 Environment Overview

• Grid Size: 5x5
• Objective: Pick up passengers and drop them at designated locations
• Challenges: Efficient navigation, avoiding unnecessary moves, maximizing reward