def main(_):
running_environment.ForceCpuForTheRun()
retro.data.merge('../../roms/contraforce.nes', quiet=False)
env = retro.make(game='ContraForce-Nes')
pipeline = shortcut.ScreenLearningPipeline(
gym_env_name='ContraForce', gym_env=env)
pipeline.Demo()
def main(_):
running_environment.ForceCpuForTheRun()
pipeline = shortcut.ScreenLearningPipeline(
gym_env_name='MsPacman-v0',
use_ddqn=False,
use_large_model=True,
)
pipeline.LoadWeights()
pipeline.Demo()
def main(_):
running_environment.ForceCpuForTheRun()
env = environment_impl.GymEnvironment(gym.make('CartPole-v0'))
brain = a3c_impl.A3C(model=a3c_impl.CreateModel(
state_shape=env.GetStateShape(),
action_space_size=env.GetActionSpaceSize(),
hidden_layer_sizes=(12, ),
))
policy = policy_impl.PolicyWithDecreasingRandomness(
base_policy=policy_impl.PiWeightedPolicy(),
initial_epsilon=0.4,
final_epsilon=0.05,
decay_by_half_after_num_of_episodes=500,
)
runner = a3c_impl.NStepExperienceRunner()
# runner = runner_impl.SimpleRunner()
runner.AddCallback(
runner_extension_impl.ProgressTracer(report_every_num_of_episodes=100))
runner.Run(env=env, brain=brain, policy=policy, num_of_episodes=1200)
"""Demo using A3C to solve CartPole-v0."""
import gym
from deep_learning.engine import a3c_impl
from deep_learning.engine import environment_impl
from deep_learning.engine import policy_impl
from deep_learning.engine import runner_extension_impl
from qpylib import running_environment
running_environment.ForceCpuForTheRun()
from absl import app
# Profiler instruction:
# 1) Generate profiler file:
# $ python -m cProfile -o result.prof deep_learning/examples/solve_cartpole_a3c.py
# 2) Visualize it:
# $ snakeviz result.prof
# It prints a link that shows the viz.
def main(_):
running_environment.ForceCpuForTheRun()
env = environment_impl.GymEnvironment(gym.make('CartPole-v0'))
brain = a3c_impl.A3C(model=a3c_impl.CreateModel(
state_shape=env.GetStateShape(),
action_space_size=env.GetActionSpaceSize(),
hidden_layer_sizes=(12, ),
))
def main(_):
running_environment.ForceCpuForTheRun()
print(gym_super_mario_bros.__doc__)
pipeline = shortcut.ScreenLearningPipeline(
gym_env_name='SuperMarioBros-v0')
pipeline.Demo()