drlkit 0.1.7

A High Level Python Deep Reinforcement Learning library. Great for beginners, for prototyping and quickly comparing algorithms

---

A High Level Python Deep Reinforcement Learning library. Great for beginners, prototyping and quickly comparing algorithms

UNDER CONSTRUCTION!

Do not use yet!

System	3.5	3.6	3.7
Linux CPU			—
Linux GPU			—
Windows CPU / GPU	—		—
Linux (ppc64le) CPU		—
Linux (ppc64le) GPU		—

Installation

Run the following to install:

pip install drlkit

Usage

import numpy as np
from drlkit import TorchAgent, Plot, EnvironmentWrapper

ENV_NAME = "LunarLander-v2"
env = EnvironmentWrapper(ENV_NAME)
agent = TorchAgent(state_size=8, action_size=env.env.action_space.n, seed=0)

# Train the agent
env.fit(agent, n_episodes=1000)

# See the results
Plot.basic_plot(np.arange(len(env.scores)), env.scores, xlabel='Episode #', ylabel='Score')


# Play untrained agent
env.load_model(agent, env="LunarLander", elapsed_episodes=3000)
env.play(num_episodes=10, trained=False)

# Play trained agent
env.play(num_episodes=10, trained=True)