.. include:: /_includes/rllib/we_are_hiring.rst .. include:: /_includes/rllib/new_api_stack.rst .. include:: /_includes/rllib/new_api_stack_component.rst .. |tensorflow| image:: images/tensorflow.png :class: inline-figure :width: 16 .. |pytorch| image:: images/pytorch.png :class: inline-figure :width: 16 .. _rlmodule-guide: RL Modules (Alpha) ================== .. note:: This is an experimental module that serves as a general replacement for ModelV2, and is subject to change. It will eventually match the functionality of the previous stack. If you only use high-level RLlib APIs such as :py:class:`~ray.rllib.algorithms.algorithm.Algorithm` you should not experience significant changes, except for a few new parameters to the configuration object. If you've used custom models or policies before, you'll need to migrate them to the new modules. Check the Migration guide for more information. The table below shows the list of migrated algorithms and their current supported features, which will be updated as we progress. .. list-table:: :header-rows: 1 :widths: 20 20 20 20 20 20 * - Algorithm - Independent MARL - Fully-connected - Image inputs (CNN) - RNN support (LSTM) - Complex observations (ComplexNet) * - **PPO** - |pytorch| |tensorflow| - |pytorch| |tensorflow| - |pytorch| |tensorflow| - - |pytorch| * - **Impala** - |pytorch| |tensorflow| - |pytorch| |tensorflow| - |pytorch| |tensorflow| - - |pytorch| * - **APPO** - |pytorch| |tensorflow| - |pytorch| |tensorflow| - |pytorch| |tensorflow| - - RL Module is a neural network container that implements three public methods: :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_train`, :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration`, and :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference`. Each method corresponds to a distinct reinforcement learning phase. :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` handles acting and data collection, balancing exploration and exploitation. On the other hand, the :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference` serves the learned model during evaluation, often being less stochastic. :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_train` manages the training phase, handling calculations exclusive to computing losses, such as learning Q values in a DQN model. Enabling RL Modules in the Configuration ---------------------------------------- Enable RL Modules via our configuration object: ``AlgorithmConfig.experimental(_enable_new_api_stack=True)``. .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __enabling-rlmodules-in-configs-begin__ :end-before: __enabling-rlmodules-in-configs-end__ Constructing RL Modules ----------------------- The RLModule API provides a unified way to define custom reinforcement learning models in RLlib. This API enables you to design and implement your own models to suit specific needs. To maintain consistency and usability, RLlib offers a standardized approach for defining module objects for both single-agent and multi-agent reinforcement learning environments. This is achieved through the :py:class:`~ray.rllib.core.rl_module.rl_module.SingleAgentRLModuleSpec` and :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModuleSpec` classes. The built-in RLModules in RLlib follow this consistent design pattern, making it easier for you to understand and utilize these modules. .. tab-set:: .. tab-item:: Single Agent .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __constructing-rlmodules-sa-begin__ :end-before: __constructing-rlmodules-sa-end__ .. tab-item:: Multi Agent .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __constructing-rlmodules-ma-begin__ :end-before: __constructing-rlmodules-ma-end__ You can pass RL Module specs to the algorithm configuration to be used by the algorithm. .. tab-set:: .. tab-item:: Single Agent .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __pass-specs-to-configs-sa-begin__ :end-before: __pass-specs-to-configs-sa-end__ .. note:: For passing RL Module specs, all fields do not have to be filled as they are filled based on the described environment or other algorithm configuration parameters (i.e. ,``observation_space``, ``action_space``, ``model_config_dict`` are not required fields when passing a custom RL Module spec to the algorithm config.) .. tab-item:: Multi Agent .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __pass-specs-to-configs-ma-begin__ :end-before: __pass-specs-to-configs-ma-end__ Writing Custom Single Agent RL Modules -------------------------------------- For single-agent algorithms (e.g., PPO, DQN) or independent multi-agent algorithms (e.g., PPO-MultiAgent), use :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule`. For more advanced multi-agent use cases with a shared communication between agents, extend the :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModule` class. RLlib treats single-agent modules as a special case of :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModule` with only one module. Create the multi-agent representation of all RLModules by calling :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.as_multi_agent`. For example: .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __convert-sa-to-ma-begin__ :end-before: __convert-sa-to-ma-end__ RLlib implements the following abstract framework specific base classes: - :class:`TorchRLModule `: For PyTorch-based RL Modules. - :class:`TfRLModule `: For TensorFlow-based RL Modules. The minimum requirement is for sub-classes of :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule` is to implement the following methods: - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule._forward_train`: Forward pass for training. - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule._forward_inference`: Forward pass for inference. - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule._forward_exploration`: Forward pass for exploration. For your custom :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` and :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference` methods, you must return a dictionary that either contains the key "actions" and/or the key "action_dist_inputs". If you return the "actions" key: - RLlib will use the actions provided thereunder as-is. - If you also returned the "action_dist_inputs" key: RLlib will also create a :py:class:`~ray.rllib.models.distributions.Distribution` object from the distribution parameters under that key and - in the case of :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` - compute action probs and logp values from the given actions automatically. If you do not return the "actions" key: - You must return the "action_dist_inputs" key instead from your :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` and :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference` methods. - RLlib will create a :py:class:`~ray.rllib.models.distributions.Distribution` object from the distribution parameters under that key and sample actions from the thus generated distribution. - In the case of :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration`, RLlib will also compute action probs and logp values from the sampled actions automatically. .. note:: In the case of :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference`, the generated distributions (from returned key "action_dist_inputs") will always be made deterministic first via the :py:meth:`~ray.rllib.models.distributions.Distribution.to_deterministic` utility before a possible action sample step. Thus, for example, sampling from a Categorical distribution will be reduced to simply selecting the argmax actions from the distribution's logits/probs. Commonly used distribution implementations can be found under ``ray.rllib.models.tf.tf_distributions`` for tensorflow and ``ray.rllib.models.torch.torch_distributions`` for torch. You can choose to return determinstic actions, by creating a determinstic distribution instance. .. tab-set:: .. tab-item:: Returning "actions" key .. code-block:: python """ An RLModule whose forward_exploration/inference methods return the "actions" key. """ class MyRLModule(TorchRLModule): ... def _forward_inference(self, batch): ... return { "actions": ... # actions will be used as-is } def _forward_exploration(self, batch): ... return { "actions": ... # actions will be used as-is (no sampling step!) "action_dist_inputs": ... # optional: If provided, will be used to compute action probs and logp. } .. tab-item:: Not returning "actions" key .. code-block:: python """ An RLModule whose forward_exploration/inference methods do NOT return the "actions" key. """ class MyRLModule(TorchRLModule): ... def _forward_inference(self, batch): ... return { # RLlib will: # - Generate distribution from these parameters. # - Convert distribution to a deterministic equivalent. # - "sample" from the deterministic distribution. "action_dist_inputs": ... } def _forward_exploration(self, batch): ... return { # RLlib will: # - Generate distribution from these parameters. # - "sample" from the (stochastic) distribution. # - Compute action probs/logs automatically using the sampled # actions and the generated distribution object. "action_dist_inputs": ... } Also the :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule` class's constrcutor requires a dataclass config object called `~ray.rllib.core.rl_module.rl_module.RLModuleConfig` which contains the following fields: - :py:attr:`~ray.rllib.core.rl_module.rl_module.RLModuleConfig.observation_space`: The observation space of the environment (either processed or raw). - :py:attr:`~ray.rllib.core.rl_module.rl_module.RLModuleConfig.action_space`: The action space of the environment. - :py:attr:`~ray.rllib.core.rl_module.rl_module.RLModuleConfig.model_config_dict`: The model config dictionary of the algorithm. Model hyper-parameters such as number of layers, type of activation, etc. are defined here. - :py:attr:`~ray.rllib.core.rl_module.rl_module.RLModuleConfig.catalog_class`: The :py:class:`~ray.rllib.core.models.catalog.Catalog` object of the algorithm. When writing RL Modules, you need to use these fields to construct your model. .. tab-set:: .. tab-item:: Single Agent (torch) .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __write-custom-sa-rlmodule-torch-begin__ :end-before: __write-custom-sa-rlmodule-torch-end__ .. tab-item:: Single Agent (tensorflow) .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __write-custom-sa-rlmodule-tf-begin__ :end-before: __write-custom-sa-rlmodule-tf-end__ In :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule` you can enforce the checking for the existence of certain input or output keys in the data that is communicated into and out of RL Modules. This serves multiple purposes: - For the I/O requirement of each method to be self-documenting. - For failures to happen quickly. If users extend the modules and implement something that does not match the assumptions of the I/O specs, the check reports missing keys and their expected format. For example, RLModule should always have an ``obs`` key in the input batch and an ``action_dist`` key in the output. .. tab-set:: .. tab-item:: Single Level Keys .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __extend-spec-checking-single-level-begin__ :end-before: __extend-spec-checking-single-level-end__ .. tab-item:: Nested Keys .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __extend-spec-checking-nested-begin__ :end-before: __extend-spec-checking-nested-end__ .. tab-item:: TensorShape Spec .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __extend-spec-checking-torch-specs-begin__ :end-before: __extend-spec-checking-torch-specs-end__ .. tab-item:: Type Spec .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __extend-spec-checking-type-specs-begin__ :end-before: __extend-spec-checking-type-specs-end__ :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule` has two methods for each forward method, totaling 6 methods that can be override to describe the specs of the input and output of each method: - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.input_specs_inference` - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.output_specs_inference` - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.input_specs_exploration` - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.output_specs_exploration` - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.input_specs_train` - :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.output_specs_train` To learn more, see the `SpecType` documentation. Writing Custom Multi-Agent RL Modules (Advanced) ------------------------------------------------ For multi-agent modules, RLlib implements :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModule`, which is a dictionary of :py:class:`~ray.rllib.core.rl_module.rl_module.RLModule` objects, one for each policy, and possibly some shared modules. The base-class implementation works for most of use cases that need to define independent neural networks for sub-groups of agents. For more complex, multi-agent use cases, where the agents share some part of their neural network, you should inherit from this class and override the default implementation. The :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModule` offers an API for constructing custom models tailored to specific needs. The key method for this customization is :py:meth:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModule`.build. The following example creates a custom multi-agent RL module with underlying modules. The modules share an encoder, which gets applied to the global part of the observations space. The local part passes through a separate encoder, specific to each policy. .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __write-custom-marlmodule-shared-enc-begin__ :end-before: __write-custom-marlmodule-shared-enc-end__ To construct this custom multi-agent RL module, pass the class to the :py:class:`~ray.rllib.core.rl_module.marl_module.MultiAgentRLModuleSpec` constructor. Also, pass the :py:class:`~ray.rllib.core.rl_module.rl_module.SingleAgentRLModuleSpec` for each agent because RLlib requires the observation, action spaces, and model hyper-parameters for each agent. .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __pass-custom-marlmodule-shared-enc-begin__ :end-before: __pass-custom-marlmodule-shared-enc-end__ Extending Existing RLlib RL Modules ----------------------------------- RLlib provides a number of RL Modules for different frameworks (e.g., PyTorch, TensorFlow, etc.). To customize existing RLModules you can change the RLModule directly by inheriting the class and changing the :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.setup` or other methods. For example, extend :py:class:`~ray.rllib.algorithms.ppo.torch.ppo_torch_rl_module.PPOTorchRLModule` and augment it with your own customization. Then pass the new customized class into the appropriate :py:class:`~ray.rllib.algorithms.algorithm_config.AlgorithmConfig`. There are two possible ways to extend existing RL Modules: .. tab-set:: .. tab-item:: Inheriting existing RL Modules The default way to extend existing RL Modules is to inherit from them and override the methods you need to customize. Then pass the new customized class into the :py:class:`~ray.rllib.algorithms.algorithm_config.AlgorithmConfig` to optimize your custom RL Module. This is the preferred approach. With it, we can define our own models explicitly within a given RL Module and don't need to interact with a Catalog, so you don't need to learn about Catalog. .. code-block:: python class MyPPORLModule(PPORLModule): def __init__(self, config: RLModuleConfig): super().__init__(config) ... # Pass in the custom RL Module class to the spec algo_config = algo_config.rl_module( rl_module_spec=SingleAgentRLModuleSpec(module_class=MyPPORLModule) ) A concrete example: If you want to replace the default encoder that RLlib builds for torch, PPO and a given observation space, you can override :py:class:`~ray.rllib.algorithms.ppo.torch.ppo_torch_rl_module.PPOTorchRLModule`'s :py:meth:`~ray.rllib.algorithms.ppo.torch.ppo_torch_rl_module.PPOTorchRLModule.__init__` to create your custom encoder instead of the default one. We do this in the following example. .. literalinclude:: ../../../rllib/examples/rl_module/mobilenet_rlm.py :language: python :start-after: __sphinx_doc_begin__ :end-before: __sphinx_doc_end__ .. tab-item:: Extending RL Module Catalog An advanced way to customize your module is by extending its :py:class:`~ray.rllib.core.models.catalog.Catalog`. The Catalog is a component that defines the default models and other sub-components for RL Modules based on factors such as ``observation_space``, ``action_space``, etc. For more information on the :py:class:`~ray.rllib.core.models.catalog.Catalog` class, refer to the `Catalog user guide `__. By modifying the Catalog, you can alter what sub-components are being built for existing RL Modules. This approach is useful mostly if you want your custom component to integrate with the decision trees that the Catalogs represent. The following use cases are examples of what may require you to extend the Catalogs: - Choosing a custom model only for a certain observation space. - Using a custom action distribution in multiple distinct Algorithms. - Reusing your custom component in many distinct RL Modules. For instance, to adapt existing ``PPORLModules`` for a custom graph observation space not supported by RLlib out-of-the-box, extend the :py:class:`~ray.rllib.core.models.catalog.Catalog` class used to create the ``PPORLModule`` and override the method responsible for returning the encoder component to ensure that your custom encoder replaces the default one initially provided by RLlib. .. code-block:: python class MyAwesomeCatalog(PPOCatalog): def build_actor_critic_encoder(): # create your awesome graph encoder here and return it pass # Pass in the custom catalog class to the spec algo_config = algo_config.rl_module( rl_module_spec=SingleAgentRLModuleSpec(catalog_class=MyAwesomeCatalog) ) Checkpointing RL Modules ------------------------ RL Modules can be checkpointed with their two methods :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.save_to_checkpoint` and :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.from_checkpoint`. The following example shows how these methods can be used outside of, or in conjunction with, an RLlib Algorithm. .. literalinclude:: doc_code/rlmodule_guide.py :language: python :start-after: __checkpointing-begin__ :end-before: __checkpointing-end__ Migrating from Custom Policies and Models to RL Modules ------------------------------------------------------- This document is for those who have implemented custom policies and models in RLlib and want to migrate to the new `~ray.rllib.core.rl_module.rl_module.RLModule` API. If you have implemented custom policies that extended the `~ray.rllib.policy.eager_tf_policy_v2.EagerTFPolicyV2` or `~ray.rllib.policy.torch_policy_v2.TorchPolicyV2` classes, you likely did so that you could either modify the behavior of constructing models and distributions (via overriding `~ray.rllib.policy.torch_policy_v2.TorchPolicyV2.make_model`, `~ray.rllib.policy.torch_policy_v2.TorchPolicyV2.make_model_and_action_dist`), control the action sampling logic (via overriding `~ray.rllib.policy.eager_tf_policy_v2.EagerTFPolicyV2.action_distribution_fn` or `~ray.rllib.policy.eager_tf_policy_v2.EagerTFPolicyV2.action_sampler_fn`), or control the logic for infernce (via overriding `~ray.rllib.policy.policy.Policy.compute_actions_from_input_dict`, `~ray.rllib.policy.policy.Policy.compute_actions`, or `~ray.rllib.policy.policy.Policy.compute_log_likelihoods`). These APIs were built with `ray.rllib.models.modelv2.ModelV2` models in mind to enable you to customize the behavior of those functions. However `~ray.rllib.core.rl_module.rl_module.RLModule` is a more general abstraction that will reduce the amount of functions that you need to override. In the new `~ray.rllib.core.rl_module.rl_module.RLModule` API the construction of the models and the action distribution class that should be used are best defined in the constructor. That RL Module is constructed automatically if users follow the instructions outlined in the sections `Enabling RL Modules in the Configuration`_ and `Constructing RL Modules`_. `~ray.rllib.policy.policy.Policy.compute_actions` and `~ray.rllib.policy.policy.Policy.compute_actions_from_input_dict` can still be used for sampling actions for inference or exploration by using the ``explore=True|False`` parameter. If called with ``explore=True`` these functions will invoke `~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` and if ``explore=False`` then they will call `~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference`. What your customization could have looked like before: .. tab-set:: .. tab-item:: ModelV2 .. code-block:: python from ray.rllib.models.torch.torch_modelv2 import TorchModelV2 from ray.rllib.policy.torch_policy_v2 import TorchPolicyV2 class MyCustomModel(TorchModelV2): """Code for your previous custom model""" ... class CustomPolicy(TorchPolicyV2): @DeveloperAPI @OverrideToImplementCustomLogic def make_model(self) -> ModelV2: """Create model. Note: only one of make_model or make_model_and_action_dist can be overridden. Returns: ModelV2 model. """ return MyCustomModel(...) .. tab-item:: ModelV2 + Distribution .. code-block:: python from ray.rllib.models.torch.torch_modelv2 import TorchModelV2 from ray.rllib.policy.torch_policy_v2 import TorchPolicyV2 class MyCustomModel(TorchModelV2): """Code for your previous custom model""" ... class CustomPolicy(TorchPolicyV2): @DeveloperAPI @OverrideToImplementCustomLogic def make_model_and_action_dist(self): """Create model and action distribution function. Returns: ModelV2 model. ActionDistribution class. """ my_model = MyCustomModel(...) # construct some ModelV2 instance here dist_class = ... # Action distribution cls return my_model, dist_class .. tab-item:: Sampler functions .. code-block:: python from ray.rllib.models.torch.torch_modelv2 import TorchModelV2 from ray.rllib.policy.torch_policy_v2 import TorchPolicyV2 class CustomPolicy(TorchPolicyV2): @DeveloperAPI @OverrideToImplementCustomLogic def action_sampler_fn( self, model: ModelV2, *, obs_batch: TensorType, state_batches: TensorType, **kwargs, ) -> Tuple[TensorType, TensorType, TensorType, List[TensorType]]: """Custom function for sampling new actions given policy. Args: model: Underlying model. obs_batch: Observation tensor batch. state_batches: Action sampling state batch. Returns: Sampled action Log-likelihood Action distribution inputs Updated state """ return None, None, None, None @DeveloperAPI @OverrideToImplementCustomLogic def action_distribution_fn( self, model: ModelV2, *, obs_batch: TensorType, state_batches: TensorType, **kwargs, ) -> Tuple[TensorType, type, List[TensorType]]: """Action distribution function for this Policy. Args: model: Underlying model. obs_batch: Observation tensor batch. state_batches: Action sampling state batch. Returns: Distribution input. ActionDistribution class. State outs. """ return None, None, None All of the ``Policy.compute_***`` functions expect that :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_exploration` and :py:meth:`~ray.rllib.core.rl_module.rl_module.RLModule.forward_inference` return a dictionary that either contains the key "actions" and/or the key "action_dist_inputs". See `Writing Custom Single Agent RL Modules`_ for more details on how to implement your own custom RL Modules. .. tab-set:: .. tab-item:: The Equivalent RL Module .. code-block:: python """ No need to override any policy functions. Simply instead implement any custom logic in your custom RL Module """ from ray.rllib.models.torch.torch_distributions import YOUR_DIST_CLASS class MyRLModule(TorchRLModule): def __init__(self, config: RLConfig): # construct any custom networks here using config # specify an action distribution class here ... def _forward_inference(self, batch): ... def _forward_exploration(self, batch): ...