Train/Eval Mode Support (Stable-Baselines-Team#39)

* switch models between train and eval mode

* update changelog

* update release in change log

* Update dependency

Co-authored-by: Antonin Raffin <antonin.raffin@ensta.org>

docs/misc/changelog.rst

@@ -4,14 +4,18 @@ Changelog
 ==========
 
 
-Release 1.2.0a0 (WIP)
+Release 1.2.0 (2021-09-08)
 -------------------------------
 
+**Train/Eval mode support**
+
 Breaking Changes:
 ^^^^^^^^^^^^^^^^^
+- Upgraded to Stable-Baselines3 >= 1.2.0
 
 Bug Fixes:
 ^^^^^^^^^^
+- QR-DQN and TQC updated so that their policies are switched between train and eval mode at the correct time (@ayeright)
 
 Deprecations:
 ^^^^^^^^^^^^^
@@ -152,4 +156,4 @@ Stable-Baselines3 is currently maintained by `Antonin Raffin`_ (aka `@araffin`_)
 Contributors:
 -------------
 
-@ku2482 @guyk1971 @minhlong94
+@ku2482 @guyk1971 @minhlong94 @ayeright

sb3_contrib/qrdqn/policies.py

@@ -167,6 +167,7 @@ def _build(self, lr_schedule: Schedule) -> None:
         self.quantile_net = self.make_quantile_net()
         self.quantile_net_target = self.make_quantile_net()
         self.quantile_net_target.load_state_dict(self.quantile_net.state_dict())
+        self.quantile_net_target.set_training_mode(False)
 
         # Setup optimizer with initial learning rate
         self.optimizer = self.optimizer_class(self.parameters(), lr=lr_schedule(1), **self.optimizer_kwargs)
@@ -199,6 +200,15 @@ def _get_constructor_parameters(self) -> Dict[str, Any]:
         )
         return data
 
+    def set_training_mode(self, mode: bool) -> None:
+        """
+        Put the policy in either training or evaluation mode.
+        This affects certain modules, such as batch normalisation and dropout.
+        :param mode: if true, set to training mode, else set to evaluation mode
+        """
+        self.quantile_net.set_training_mode(mode)
+        self.training = mode
+
 
 MlpPolicy = QRDQNPolicy
 

sb3_contrib/qrdqn/qrdqn.py

@@ -155,6 +155,8 @@ def _on_step(self) -> None:
         self.logger.record("rollout/exploration rate", self.exploration_rate)
 
     def train(self, gradient_steps: int, batch_size: int = 100) -> None:
+        # Switch to train mode (this affects batch norm / dropout)
+        self.policy.set_training_mode(True)
         # Update learning rate according to schedule
         self._update_learning_rate(self.policy.optimizer)
 

sb3_contrib/tqc/policies.py

@@ -376,6 +376,9 @@ def _build(self, lr_schedule: Schedule) -> None:
         self.critic_target = self.make_critic(features_extractor=None)
         self.critic_target.load_state_dict(self.critic.state_dict())
 
+        # Target networks should always be in eval mode
+        self.critic_target.set_training_mode(False)
+
         self.critic.optimizer = self.optimizer_class(critic_parameters, lr=lr_schedule(1), **self.optimizer_kwargs)
 
     def _get_constructor_parameters(self) -> Dict[str, Any]:
@@ -423,6 +426,16 @@ def forward(self, obs: th.Tensor, deterministic: bool = False) -> th.Tensor:
     def _predict(self, observation: th.Tensor, deterministic: bool = False) -> th.Tensor:
         return self.actor(observation, deterministic)
 
+    def set_training_mode(self, mode: bool) -> None:
+        """
+        Put the policy in either training or evaluation mode.
+        This affects certain modules, such as batch normalisation and dropout.
+        :param mode: if true, set to training mode, else set to evaluation mode
+        """
+        self.actor.set_training_mode(mode)
+        self.critic.set_training_mode(mode)
+        self.training = mode
+
 
 MlpPolicy = TQCPolicy
 

sb3_contrib/tqc/tqc.py

@@ -175,6 +175,8 @@ def _create_aliases(self) -> None:
         self.critic_target = self.policy.critic_target
 
     def train(self, gradient_steps: int, batch_size: int = 64) -> None:
+        # Switch to train mode (this affects batch norm / dropout)
+        self.policy.set_training_mode(True)
         # Update optimizers learning rate
         optimizers = [self.actor.optimizer, self.critic.optimizer]
         if self.ent_coef_optimizer is not None:

sb3_contrib/version.txt

@@ -1 +1 @@
-1.2.0a0
+1.2.0

setup.py

@@ -62,7 +62,7 @@
     packages=[package for package in find_packages() if package.startswith("sb3_contrib")],
     package_data={"sb3_contrib": ["py.typed", "version.txt"]},
     install_requires=[
-        "stable_baselines3>=1.1.0",
+        "stable_baselines3>=1.2.0",
     ],
     description="Contrib package of Stable Baselines3, experimental code.",
     author="Antonin Raffin",

tests/test_train_eval_mode.py

@@ -0,0 +1,225 @@
+import gym
+import numpy as np
+import pytest
+import torch as th
+import torch.nn as nn
+from stable_baselines3.common.preprocessing import get_flattened_obs_dim
+from stable_baselines3.common.torch_layers import BaseFeaturesExtractor
+
+from sb3_contrib import QRDQN, TQC
+
+
+class FlattenBatchNormDropoutExtractor(BaseFeaturesExtractor):
+    """
+    Feature extract that flatten the input and applies batch normalization and dropout.
+    Used as a placeholder when feature extraction is not needed.
+    :param observation_space:
+    """
+
+    def __init__(self, observation_space: gym.Space):
+        super(FlattenBatchNormDropoutExtractor, self).__init__(
+            observation_space,
+            get_flattened_obs_dim(observation_space),
+        )
+        self.flatten = nn.Flatten()
+        self.batch_norm = nn.BatchNorm1d(self._features_dim)
+        self.dropout = nn.Dropout(0.5)
+
+    def forward(self, observations: th.Tensor) -> th.Tensor:
+        result = self.flatten(observations)
+        result = self.batch_norm(result)
+        result = self.dropout(result)
+        return result
+
+
+def clone_batch_norm_stats(batch_norm: nn.BatchNorm1d) -> (th.Tensor, th.Tensor):
+    """
+    Clone the bias and running mean from the given batch norm layer.
+    :param batch_norm:
+    :return: the bias and running mean
+    """
+    return batch_norm.bias.clone(), batch_norm.running_mean.clone()
+
+
+def clone_qrdqn_batch_norm_stats(model: QRDQN) -> (th.Tensor, th.Tensor, th.Tensor, th.Tensor):
+    """
+    Clone the bias and running mean from the quantile network and quantile-target network.
+    :param model:
+    :return: the bias and running mean from the quantile network and quantile-target network
+    """
+    quantile_net_batch_norm = model.policy.quantile_net.features_extractor.batch_norm
+    quantile_net_bias, quantile_net_running_mean = clone_batch_norm_stats(quantile_net_batch_norm)
+
+    quantile_net_target_batch_norm = model.policy.quantile_net_target.features_extractor.batch_norm
+    quantile_net_target_bias, quantile_net_target_running_mean = clone_batch_norm_stats(quantile_net_target_batch_norm)
+
+    return quantile_net_bias, quantile_net_running_mean, quantile_net_target_bias, quantile_net_target_running_mean
+
+
+def clone_tqc_batch_norm_stats(
+    model: TQC,
+) -> (th.Tensor, th.Tensor, th.Tensor, th.Tensor, th.Tensor, th.Tensor):
+    """
+    Clone the bias and running mean from the actor and critic networks and critic-target networks.
+    :param model:
+    :return: the bias and running mean from the actor and critic networks and critic-target networks
+    """
+    actor_batch_norm = model.actor.features_extractor.batch_norm
+    actor_bias, actor_running_mean = clone_batch_norm_stats(actor_batch_norm)
+
+    critic_batch_norm = model.critic.features_extractor.batch_norm
+    critic_bias, critic_running_mean = clone_batch_norm_stats(critic_batch_norm)
+
+    critic_target_batch_norm = model.critic_target.features_extractor.batch_norm
+    critic_target_bias, critic_target_running_mean = clone_batch_norm_stats(critic_target_batch_norm)
+
+    return (actor_bias, actor_running_mean, critic_bias, critic_running_mean, critic_target_bias, critic_target_running_mean)
+
+
+CLONE_HELPERS = {
+    QRDQN: clone_qrdqn_batch_norm_stats,
+    TQC: clone_tqc_batch_norm_stats,
+}
+
+
+def test_qrdqn_train_with_batch_norm():
+    model = QRDQN(
+        "MlpPolicy",
+        "CartPole-v1",
+        policy_kwargs=dict(net_arch=[16, 16], features_extractor_class=FlattenBatchNormDropoutExtractor),
+        learning_starts=0,
+        seed=1,
+        tau=0,  # do not clone the target
+    )
+
+    (
+        quantile_net_bias_before,
+        quantile_net_running_mean_before,
+        quantile_net_target_bias_before,
+        quantile_net_target_running_mean_before,
+    ) = clone_qrdqn_batch_norm_stats(model)
+
+    model.learn(total_timesteps=200)
+
+    (
+        quantile_net_bias_after,
+        quantile_net_running_mean_after,
+        quantile_net_target_bias_after,
+        quantile_net_target_running_mean_after,
+    ) = clone_qrdqn_batch_norm_stats(model)
+
+    assert ~th.isclose(quantile_net_bias_before, quantile_net_bias_after).all()
+    assert ~th.isclose(quantile_net_running_mean_before, quantile_net_running_mean_after).all()
+
+    assert th.isclose(quantile_net_target_bias_before, quantile_net_target_bias_after).all()
+    assert th.isclose(quantile_net_target_running_mean_before, quantile_net_target_running_mean_after).all()
+
+
+def test_tqc_train_with_batch_norm():
+    model = TQC(
+        "MlpPolicy",
+        "Pendulum-v0",
+        policy_kwargs=dict(net_arch=[16, 16], features_extractor_class=FlattenBatchNormDropoutExtractor),
+        learning_starts=0,
+        tau=0,  # do not copy the target
+        seed=1,
+    )
+
+    (
+        actor_bias_before,
+        actor_running_mean_before,
+        critic_bias_before,
+        critic_running_mean_before,
+        critic_target_bias_before,
+        critic_target_running_mean_before,
+    ) = clone_tqc_batch_norm_stats(model)
+
+    model.learn(total_timesteps=200)
+
+    (
+        actor_bias_after,
+        actor_running_mean_after,
+        critic_bias_after,
+        critic_running_mean_after,
+        critic_target_bias_after,
+        critic_target_running_mean_after,
+    ) = clone_tqc_batch_norm_stats(model)
+
+    assert ~th.isclose(actor_bias_before, actor_bias_after).all()
+    assert ~th.isclose(actor_running_mean_before, actor_running_mean_after).all()
+
+    assert ~th.isclose(critic_bias_before, critic_bias_after).all()
+    assert ~th.isclose(critic_running_mean_before, critic_running_mean_after).all()
+
+    assert th.isclose(critic_target_bias_before, critic_target_bias_after).all()
+    assert th.isclose(critic_target_running_mean_before, critic_target_running_mean_after).all()
+
+
+@pytest.mark.parametrize("model_class", [QRDQN, TQC])
+def test_offpolicy_collect_rollout_batch_norm(model_class):
+    if model_class in [QRDQN]:
+        env_id = "CartPole-v1"
+    else:
+        env_id = "Pendulum-v0"
+
+    clone_helper = CLONE_HELPERS[model_class]
+
+    learning_starts = 10
+    model = model_class(
+        "MlpPolicy",
+        env_id,
+        policy_kwargs=dict(net_arch=[16, 16], features_extractor_class=FlattenBatchNormDropoutExtractor),
+        learning_starts=learning_starts,
+        seed=1,
+        gradient_steps=0,
+        train_freq=1,
+    )
+
+    batch_norm_stats_before = clone_helper(model)
+
+    model.learn(total_timesteps=100)
+
+    batch_norm_stats_after = clone_helper(model)
+
+    # No change in batch norm params
+    for param_before, param_after in zip(batch_norm_stats_before, batch_norm_stats_after):
+        assert th.isclose(param_before, param_after).all()
+
+
+@pytest.mark.parametrize("model_class", [QRDQN, TQC])
+@pytest.mark.parametrize("env_id", ["Pendulum-v0", "CartPole-v1"])
+def test_predict_with_dropout_batch_norm(model_class, env_id):
+    if env_id == "CartPole-v1":
+        if model_class in [TQC]:
+            return
+    elif model_class in [QRDQN]:
+        return
+
+    model_kwargs = dict(seed=1)
+    clone_helper = CLONE_HELPERS[model_class]
+
+    if model_class in [QRDQN, TQC]:
+        model_kwargs["learning_starts"] = 0
+    else:
+        model_kwargs["n_steps"] = 64
+
+    policy_kwargs = dict(
+        features_extractor_class=FlattenBatchNormDropoutExtractor,
+        net_arch=[16, 16],
+    )
+    model = model_class("MlpPolicy", env_id, policy_kwargs=policy_kwargs, verbose=1, **model_kwargs)
+
+    batch_norm_stats_before = clone_helper(model)
+
+    env = model.get_env()
+    observation = env.reset()
+    first_prediction, _ = model.predict(observation, deterministic=True)
+    for _ in range(5):
+        prediction, _ = model.predict(observation, deterministic=True)
+        np.testing.assert_allclose(first_prediction, prediction)
+
+    batch_norm_stats_after = clone_helper(model)
+
+    # No change in batch norm params
+    for param_before, param_after in zip(batch_norm_stats_before, batch_norm_stats_after):
+        assert th.isclose(param_before, param_after).all()