Q-Learning Agent playing FrozenLake-v1
This is a trained Q-Learning agent playing FrozenLake-v1.
Usage
import gymnasium as gym
from huggingface_hub import snapshot_download
# https://github.com/libertininick/r2seedo
from r2seedo.io import load_n_verify_model
# Download model snapshot from Hugging Face Hub
repo_local_path = snapshot_download(
repo_id="libertininick/q-table-frozen-lake",
local_dir="path/to/download",
)
# Load the model from the snapshot
model = load_n_verify_model(repo_local_path)
# Create the environment
env = env_slippery = gym.make(
id='FrozenLake-v1',
map_name='4x4',
is_slippery=True,
)
- Downloads last month
- 0
Evaluation results
- mean_reward on FrozenLake-v1-4x4-slipperyself-reported0.75 +/- 0.43