Init Algorithm Module (#58)

Author: pan-x-c

trinity/algorithm/__init__.py

@@ -0,0 +1,9 @@
+from trinity.algorithm.advantage_fn.advantage_fn import ADVANTAGE_FN, AdvantageFn
+from trinity.algorithm.policy_loss_fn.policy_loss_fn import POLICY_LOSS_FN, PolicyLossFn
+
+__all__ = [
+    "AdvantageFn",
+    "ADVANTAGE_FN",
+    "PolicyLossFn",
+    "POLICY_LOSS_FN",
+]

trinity/algorithm/advantage_fn/__init__.py

@@ -0,0 +1,21 @@
+from abc import ABC, abstractmethod
+from typing import Any, Dict, Tuple
+
+from trinity.utils.registry import Registry
+
+ADVANTAGE_FN = Registry("advantage_fn")
+
+
+class AdvantageFn(ABC):
+    @abstractmethod
+    def __call__(self, exps: Any, **kwargs: Dict) -> Tuple[Any, Dict]:
+        """Calculate advantages from experiences
+
+        Args:
+            exps (`DataProto`): The input experiences.
+            kwargs (`Dict`): The step-level parameters for calculating advantages.
+
+        Returns:
+            `Any`: The experiences with advantages.
+            `Dict`: The metrics for logging.
+        """

trinity/algorithm/advantage_fn/advantage_fn.py

@@ -0,0 +1,38 @@
+from abc import ABC, abstractmethod
+from typing import Any, Dict, Tuple
+
+import torch
+
+from trinity.utils.registry import Registry
+
+POLICY_LOSS_FN = Registry("policy_loss_fn")
+
+
+class PolicyLossFn(ABC):
+    """
+    Policy Loss Function
+    """
+
+    @abstractmethod
+    def __call__(
+        self,
+        logprob: torch.Tensor,
+        old_logprob: torch.Tensor,
+        action_mask: torch.Tensor,
+        advantages: torch.Tensor,
+        experiences: Any,
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Dict]:
+        """
+        Args:
+            logprob (`torch.Tensor`): The log probability generated by the policy model.
+            old_logprob (`torch.Tensor`): The log probability generated by the reference model.
+            action_mask (`torch.Tensor`): The action mask.
+            advantages (`torch.Tensor`): The advantages.
+            experiences (`DataProto`): The input experiences.
+            kwargs (`Dict`): The step-level parameters for calculating the policy loss.
+
+        Returns:
+            `torch.Tensor`: Policy loss
+            `Dict`: The metrics for logging.
+        """