def setUp(self) -> None:
self.env = ToTensor(gym.make("CartPole-v0"))
self.obs_shape = self.env.observation_space.shape
self.n_actions = self.env.action_space.n
self.net = MLP(self.obs_shape, self.n_actions)
self.agent = Agent(self.net)
self.xp_stream = EpisodicExperienceStream(self.env,
self.agent,
Mock(),
episodes=4)
self.rl_dataloader = DataLoader(self.xp_stream)
parent_parser = argparse.ArgumentParser(add_help=False)
parent_parser = cli.add_base_args(parent=parent_parser)
parent_parser = DQN.add_model_specific_args(parent_parser)
args_list = [
"--algo",
"dqn",
"--warm_start_steps",
"500",
"--episode_length",
"100",
]
self.hparams = parent_parser.parse_args(args_list)
self.model = Reinforce(**vars(self.hparams))
def setUp(self) -> None:
parent_parser = argparse.ArgumentParser(add_help=False)
parent_parser = pl.Trainer.add_argparse_args(parent_parser)
parent_parser = cli.add_base_args(parent=parent_parser)
parent_parser = DQN.add_model_specific_args(parent_parser)
args_list = [
"--algo",
"dqn",
"--n_steps",
"4",
"--warm_start_steps",
"100",
"--episode_length",
"100",
"--gpus",
"0",
"--env",
"PongNoFrameskip-v4",
]
self.hparams = parent_parser.parse_args(args_list)
self.trainer = pl.Trainer(
gpus=self.hparams.gpus,
max_steps=100,
max_epochs=
100, # Set this as the same as max steps to ensure that it doesn't stop early
val_check_interval=
1, # This just needs 'some' value, does not effect training right now
fast_dev_run=True)
def cli_main():
parser = argparse.ArgumentParser(add_help=False)
# trainer args
parser = pl.Trainer.add_argparse_args(parser)
# model args
parser = cli.add_base_args(parser)
parser = DQN.add_model_specific_args(parser)
args = parser.parse_args()
model = DQN(**args.__dict__)
trainer = pl.Trainer.from_argparse_args(args)
trainer.fit(model)
def setUp(self) -> None:
self.env = ToTensor(gym.make("CartPole-v0"))
self.obs_shape = self.env.observation_space.shape
self.n_actions = self.env.action_space.n
self.net = MLP(self.obs_shape, self.n_actions)
self.agent = Agent(self.net)
parent_parser = argparse.ArgumentParser(add_help=False)
parent_parser = cli.add_base_args(parent=parent_parser)
parent_parser = PolicyGradient.add_model_specific_args(parent_parser)
args_list = [
"--episode_length",
"100",
"--env",
"CartPole-v0",
]
self.hparams = parent_parser.parse_args(args_list)
self.model = PolicyGradient(**vars(self.hparams))
def setUp(self) -> None:
parent_parser = argparse.ArgumentParser(add_help=False)
parent_parser = cli.add_base_args(parent=parent_parser)
parent_parser = PolicyGradient.add_model_specific_args(parent_parser)
args_list = [
"--algo", "PolicyGradient", "--episode_length", "100", "--env",
"CartPole-v0"
]
self.hparams = parent_parser.parse_args(args_list)
self.trainer = pl.Trainer(
gpus=0,
max_steps=100,
max_epochs=
100, # Set this as the same as max steps to ensure that it doesn't stop early
val_check_interval=
1000, # This just needs 'some' value, does not effect training right now
fast_dev_run=True)
def setUp(self) -> None:
self.env = ToTensor(gym.make("CartPole-v0"))
self.obs_shape = self.env.observation_space.shape
self.n_actions = self.env.action_space.n
self.net = MLP(self.obs_shape, self.n_actions)
self.agent = Agent(self.net)
self.exp_source = DiscountedExperienceSource(self.env, self.agent)
parent_parser = argparse.ArgumentParser(add_help=False)
parent_parser = cli.add_base_args(parent=parent_parser)
parent_parser = DQN.add_model_specific_args(parent_parser)
args_list = [
"--algo", "dqn", "--warm_start_steps", "500", "--episode_length",
"100", "--env", "CartPole-v0", "--batch_size", "32", "--gamma",
"0.99"
]
self.hparams = parent_parser.parse_args(args_list)
self.model = Reinforce(**vars(self.hparams))
self.rl_dataloader = self.model.train_dataloader()
def cli_main():
parser = argparse.ArgumentParser(add_help=False)
# trainer args
parser = pl.Trainer.add_argparse_args(parser)
# model args
parser = cli.add_base_args(parser)
parser = VanillaPolicyGradient.add_model_specific_args(parser)
args = parser.parse_args()
model = VanillaPolicyGradient(**args.__dict__)
# save checkpoints based on avg_reward
checkpoint_callback = ModelCheckpoint(
save_top_k=1, monitor="avg_reward", mode="max", period=1, verbose=True
)
seed_everything(123)
trainer = pl.Trainer.from_argparse_args(
args, deterministic=True, checkpoint_callback=checkpoint_callback
)
trainer.fit(model)
"episodes": self.episode_count,
"episode_steps": self.episode_steps,
"epsilon": self.agent.epsilon,
}
return OrderedDict({
"loss": loss,
"avg_reward": self.avg_reward,
"log": log,
"progress_bar": status,
})
# todo: covert to CLI func and add test
if __name__ == '__main__':
parser = argparse.ArgumentParser(add_help=False)
# trainer args
parser = pl.Trainer.add_argparse_args(parser)
# model args
parser = cli.add_base_args(parser)
parser = DoubleDQN.add_model_specific_args(parser)
args = parser.parse_args()
model = DoubleDQN(**args.__dict__)
trainer = pl.Trainer.from_argparse_args(args)
trainer.fit(model)
