mirror of
https://github.com/vale981/ray
synced 2025-03-06 10:31:39 -05:00
60d4d5e1aa
* Remove all __future__ imports from RLlib. * Remove (object) again from tf_run_builder.py::TFRunBuilder. * Fix 2xLINT warnings. * Fix broken appo_policy import (must be appo_tf_policy) * Remove future imports from all other ray files (not just RLlib). * Remove future imports from all other ray files (not just RLlib). * Remove future import blocks that contain `unicode_literals` as well. Revert appo_tf_policy.py to appo_policy.py (belongs to another PR). * Add two empty lines before Schedule class. * Put back __future__ imports into determine_tests_to_run.py. Fails otherwise on a py2/print related error.
53 lines
1.5 KiB
Python
53 lines
1.5 KiB
Python
import argparse
|
|
|
|
import ray
|
|
from ray import tune
|
|
from ray.rllib.agents.trainer_template import build_trainer
|
|
from ray.rllib.evaluation.postprocessing import discount
|
|
from ray.rllib.policy.tf_policy_template import build_tf_policy
|
|
from ray.rllib.utils import try_import_tf
|
|
|
|
tf = try_import_tf()
|
|
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("--iters", type=int, default=200)
|
|
|
|
|
|
def policy_gradient_loss(policy, model, dist_class, train_batch):
|
|
logits, _ = model.from_batch(train_batch)
|
|
action_dist = dist_class(logits, model)
|
|
return -tf.reduce_mean(
|
|
action_dist.logp(train_batch["actions"]) * train_batch["returns"])
|
|
|
|
|
|
def calculate_advantages(policy,
|
|
sample_batch,
|
|
other_agent_batches=None,
|
|
episode=None):
|
|
sample_batch["returns"] = discount(sample_batch["rewards"], 0.99)
|
|
return sample_batch
|
|
|
|
|
|
# <class 'ray.rllib.policy.tf_policy_template.MyTFPolicy'>
|
|
MyTFPolicy = build_tf_policy(
|
|
name="MyTFPolicy",
|
|
loss_fn=policy_gradient_loss,
|
|
postprocess_fn=calculate_advantages,
|
|
)
|
|
|
|
# <class 'ray.rllib.agents.trainer_template.MyCustomTrainer'>
|
|
MyTrainer = build_trainer(
|
|
name="MyCustomTrainer",
|
|
default_policy=MyTFPolicy,
|
|
)
|
|
|
|
if __name__ == "__main__":
|
|
ray.init()
|
|
args = parser.parse_args()
|
|
tune.run(
|
|
MyTrainer,
|
|
stop={"training_iteration": args.iters},
|
|
config={
|
|
"env": "CartPole-v0",
|
|
"num_workers": 2,
|
|
})
|