added graders file

__init__.py

@@ -8,9 +8,13 @@
 
 from .client import QuantumOpenenvEnv
 from .models import QuantumAction, QuantumObservation
+from .graders import grade_easy, grade_medium, grade_hard
 
 __all__ = [
     "QuantumAction",
     "QuantumObservation",
     "QuantumOpenenvEnv",
+    "grade_easy",
+    "grade_medium",
+    "grade_hard",
 ]

graders.py

@@ -0,0 +1,44 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""
+Standalone graders for the Quantum Circuit Optimization Environment.
+"""
+
+from .models import QuantumObservation
+
+
+def grade_easy(observation: QuantumObservation) -> float:
+    """Independent grader for Easy Task."""
+    final_count = observation.gate_count
+    metadata = observation.metadata or {}
+    initial_count = metadata.get("initial_count", final_count)
+    if initial_count == 0:
+        return 1.0
+    compression_ratio = (initial_count - final_count) / initial_count
+    return max(0.0, min(1.0, compression_ratio))
+
+
+def grade_medium(observation: QuantumObservation) -> float:
+    """Independent grader for Medium Task."""
+    final_count = observation.gate_count
+    metadata = observation.metadata or {}
+    initial_count = metadata.get("initial_count", final_count)
+    if initial_count == 0:
+        return 1.0
+    compression_ratio = (initial_count - final_count) / initial_count
+    return max(0.0, min(1.0, compression_ratio))
+
+
+def grade_hard(observation: QuantumObservation) -> float:
+    """Independent grader for Hard Task."""
+    final_count = observation.gate_count
+    metadata = observation.metadata or {}
+    initial_count = metadata.get("initial_count", final_count)
+    if initial_count == 0:
+        return 1.0
+    compression_ratio = (initial_count - final_count) / initial_count
+    return max(0.0, min(1.0, compression_ratio))

openenv.yaml

@@ -1,22 +1,22 @@
 name: "quantum_circuit_optimizer"
 version: "0.1.0"
 description: "An OpenEnv environment for optimizing multi-qubit quantum circuits using RL/LLM agents."
-entrypoint: "server.quantum_openenv_env_environment:QuantumCircuitOptimizationEnvironment"
+entrypoint: "quantum_openenv_env.server.quantum_openenv_env_environment:QuantumCircuitOptimizationEnvironment"
 tasks:
   - id: "easy"
     name: "Easy Circuit Optimization"
     description: "2 qubits, single-qubit gates only."
     difficulty: "easy"
-    grader: "server.quantum_openenv_env_environment:grade_easy"
+    grader: "quantum_openenv_env.server.quantum_openenv_env_environment:QuantumCircuitOptimizationEnvironment.grade_easy"
     
   - id: "medium"
     name: "Medium Circuit Optimization"
     description: "4 qubits with basic CNOT/SWAP entanglement."
     difficulty: "medium"
-    grader: "server.quantum_openenv_env_environment:grade_medium"
+    grader: "quantum_openenv_env.server.quantum_openenv_env_environment:QuantumCircuitOptimizationEnvironment.grade_medium"
     
   - id: "hard"
     name: "Hard Circuit Optimization"
     description: "6 qubits with deep entanglement and complex blockades."
     difficulty: "hard"
-    grader: "server.quantum_openenv_env_environment:grade_hard"
+    grader: "quantum_openenv_env.server.quantum_openenv_env_environment:QuantumCircuitOptimizationEnvironment.grade_hard"

server/__init__.py

@@ -6,12 +6,8 @@
 
 """Quantum Openenv Env environment server components."""
 
-from .quantum_openenv_env_environment import (
-    QuantumCircuitOptimizationEnvironment,
-    grade_easy,
-    grade_medium,
-    grade_hard,
-)
+from .quantum_openenv_env_environment import QuantumCircuitOptimizationEnvironment
+from quantum_openenv_env.graders import grade_easy, grade_medium, grade_hard
 
 __all__ = [
     "QuantumCircuitOptimizationEnvironment",

server/quantum_openenv_env_environment.py

@@ -80,37 +80,11 @@ TASK_CONFIGS = {
 # Graders
 # ============================================================================
 
-def grade_easy(observation: QuantumObservation) -> float:
-    """Independent grader for Easy Task."""
-    final_count = observation.gate_count
-    metadata = observation.metadata or {}
-    initial_count = metadata.get("initial_count", final_count)
-    if initial_count == 0:
-        return 1.0
-    compression_ratio = (initial_count - final_count) / initial_count
-    return max(0.0, min(1.0, compression_ratio))
-
-def grade_medium(observation: QuantumObservation) -> float:
-    """Independent grader for Medium Task."""
-    final_count = observation.gate_count
-    metadata = observation.metadata or {}
-    initial_count = metadata.get("initial_count", final_count)
-    if initial_count == 0:
-        return 1.0
-    compression_ratio = (initial_count - final_count) / initial_count
-    return max(0.0, min(1.0, compression_ratio))
-
-def grade_hard(observation: QuantumObservation) -> float:
-    """Independent grader for Hard Task."""
-    final_count = observation.gate_count
-    metadata = observation.metadata or {}
-    initial_count = metadata.get("initial_count", final_count)
-    if initial_count == 0:
-        return 1.0
-    compression_ratio = (initial_count - final_count) / initial_count
-    return max(0.0, min(1.0, compression_ratio))
+# Graders are now defined in ../graders.py
 
 # Exporting for inference.py and Hackathon Platform
+from quantum_openenv_env.graders import grade_easy, grade_medium, grade_hard
+
 GRADERS = {
     "easy": grade_easy,
     "medium": grade_medium,
@@ -258,6 +232,19 @@ class QuantumCircuitOptimizationEnvironment(Environment):
 
         return self._build_observation(reward, action_result)
 
+    # Grader methods for OpenEnv validation
+    def grade_easy(self, observation: QuantumObservation) -> float:
+        """Grader for Easy Task."""
+        return grade_easy(observation)
+
+    def grade_medium(self, observation: QuantumObservation) -> float:
+        """Grader for Medium Task."""
+        return grade_medium(observation)
+
+    def grade_hard(self, observation: QuantumObservation) -> float:
+        """Grader for Hard Task."""
+        return grade_hard(observation)
+
     def _build_observation(self, reward: float, action_result: str) -> QuantumObservation:
         max_steps_reached = self._state.step_count >= 150
         is_done = max_steps_reached or self._is_circuit_dead_end()