ray/rllib/examples/custom_keras_model.py

"""Example of using a custom ModelV2 Keras-style model."""

import argparse
import os

import ray
from ray import tune
from ray.rllib.agents.dqn.distributional_q_tf_model import \
    DistributionalQTFModel
from ray.rllib.models import ModelCatalog
from ray.rllib.models.tf.misc import normc_initializer
from ray.rllib.models.tf.tf_modelv2 import TFModelV2
from ray.rllib.models.tf.visionnet import VisionNetwork as MyVisionNetwork
from ray.rllib.policy.policy import LEARNER_STATS_KEY
from ray.rllib.policy.sample_batch import DEFAULT_POLICY_ID
from ray.rllib.utils.framework import try_import_tf

tf1, tf, tfv = try_import_tf()

parser = argparse.ArgumentParser()
parser.add_argument(
    "--run",
    type=str,
    default="DQN",
    help="The RLlib-registered algorithm to use.")
parser.add_argument("--stop", type=int, default=200)
parser.add_argument("--use-vision-network", action="store_true")
parser.add_argument("--num-cpus", type=int, default=0)


class MyKerasModel(TFModelV2):
    """Custom model for policy gradient algorithms."""

    def __init__(self, obs_space, action_space, num_outputs, model_config,
                 name):
        super(MyKerasModel, self).__init__(obs_space, action_space,
                                           num_outputs, model_config, name)
        self.inputs = tf.keras.layers.Input(
            shape=obs_space.shape, name="observations")
        layer_1 = tf.keras.layers.Dense(
            128,
            name="my_layer1",
            activation=tf.nn.relu,
            kernel_initializer=normc_initializer(1.0))(self.inputs)
        layer_out = tf.keras.layers.Dense(
            num_outputs,
            name="my_out",
            activation=None,
            kernel_initializer=normc_initializer(0.01))(layer_1)
        value_out = tf.keras.layers.Dense(
            1,
            name="value_out",
            activation=None,
            kernel_initializer=normc_initializer(0.01))(layer_1)
        self.base_model = tf.keras.Model(self.inputs, [layer_out, value_out])

    def forward(self, input_dict, state, seq_lens):
        model_out, self._value_out = self.base_model(input_dict["obs"])
        return model_out, state

    def value_function(self):
        return tf.reshape(self._value_out, [-1])

    def metrics(self):
        return {"foo": tf.constant(42.0)}


class MyKerasQModel(DistributionalQTFModel):
    """Custom model for DQN."""

    def __init__(self, obs_space, action_space, num_outputs, model_config,
                 name, **kw):
        super(MyKerasQModel, self).__init__(
            obs_space, action_space, num_outputs, model_config, name, **kw)

        # Define the core model layers which will be used by the other
        # output heads of DistributionalQModel
        self.inputs = tf.keras.layers.Input(
            shape=obs_space.shape, name="observations")
        layer_1 = tf.keras.layers.Dense(
            128,
            name="my_layer1",
            activation=tf.nn.relu,
            kernel_initializer=normc_initializer(1.0))(self.inputs)
        layer_out = tf.keras.layers.Dense(
            num_outputs,
            name="my_out",
            activation=tf.nn.relu,
            kernel_initializer=normc_initializer(1.0))(layer_1)
        self.base_model = tf.keras.Model(self.inputs, layer_out)

    # Implement the core forward method.
    def forward(self, input_dict, state, seq_lens):
        model_out = self.base_model(input_dict["obs"])
        return model_out, state

    def metrics(self):
        return {"foo": tf.constant(42.0)}


if __name__ == "__main__":
    args = parser.parse_args()
    ray.init(num_cpus=args.num_cpus or None)
    ModelCatalog.register_custom_model(
        "keras_model", MyVisionNetwork
        if args.use_vision_network else MyKerasModel)
    ModelCatalog.register_custom_model(
        "keras_q_model", MyVisionNetwork
        if args.use_vision_network else MyKerasQModel)

    # Tests https://github.com/ray-project/ray/issues/7293
    def check_has_custom_metric(result):
        r = result["result"]["info"]["learner"]
        if DEFAULT_POLICY_ID in r:
            r = r[DEFAULT_POLICY_ID].get(LEARNER_STATS_KEY,
                                         r[DEFAULT_POLICY_ID])
        assert r["model"]["foo"] == 42, result

    if args.run == "DQN":
        extra_config = {"learning_starts": 0}
    else:
        extra_config = {}

    tune.run(
        args.run,
        stop={"episode_reward_mean": args.stop},
        config=dict(
            extra_config,
            **{
                "env": "BreakoutNoFrameskip-v4"
                if args.use_vision_network else "CartPole-v0",
                # Use GPUs iff `RLLIB_NUM_GPUS` env var set to > 0.
                "num_gpus": int(os.environ.get("RLLIB_NUM_GPUS", "0")),
                "callbacks": {
                    "on_train_result": check_has_custom_metric,
                },
                "model": {
                    "custom_model": "keras_q_model"
                    if args.run == "DQN" else "keras_model"
                },
                "framework": "tf",
            }))