inference

inference.py

@@ -16,12 +16,15 @@ from AntiAtropos.models import ActionType, SREAction
 
 load_dotenv()
 
-API_BASE_URL = os.getenv("API_BASE_URL", "https://api.openai.com/v1")
+API_BASE_URL = os.getenv("API_BASE_URL", "https://router.huggingface.co/v1")
 MODEL_NAME = os.getenv("MODEL_NAME", "gpt-4.1-mini")
-API_KEY = os.getenv("HF_TOKEN") or os.getenv("API_KEY")
+API_KEY = os.getenv("API_KEY")
+if not API_KEY:
+    # Local fallback to keep developer runs convenient.
+    API_KEY = os.getenv("HF_TOKEN") or os.getenv("OPENAI_API_KEY")
 LOCAL_IMAGE_NAME = os.getenv("LOCAL_IMAGE_NAME")
 
-ENV_URL = os.getenv("ANTIATROPOS_ENV_URL", "https://pranavkk-antiatropos.hf.space")
+ENV_URL = os.getenv("ENV_URL") or os.getenv("ANTIATROPOS_ENV_URL", "https://pranavkk-antiatropos.hf.space")
 ENV_MODE = os.getenv("ANTIATROPOS_MODE", "simulated")
 TASK_NAME = os.getenv("ANTIATROPOS_TASK", "task-1")
 BENCHMARK = os.getenv("ANTIATROPOS_BENCHMARK", "antiatropos")
@@ -67,10 +70,10 @@ def log_step(step: int, action: str, reward: float, done: bool, error: Optional[
     )
 
 
-def log_end(success: bool, steps: int, score: float, rewards: List[float]) -> None:
+def log_end(task: str, success: bool, steps: int, score: float, rewards: List[float]) -> None:
     rewards_str = ",".join(f"{r:.2f}" for r in rewards)
     print(
-        f"[END] success={str(success).lower()} steps={steps} score={score:.2f} rewards={rewards_str}",
+        f"[END] task={task} success={str(success).lower()} steps={steps} score={score:.3f} rewards={rewards_str}",
         flush=True,
     )
 
@@ -90,6 +93,10 @@ def _task_seed(base_seed: int, task_id: str) -> int:
     return int(base_seed + offsets.get(task_id, 0))
 
 
+def _strict_score(score: float, eps: float = 0.001) -> float:
+    return min(1.0 - eps, max(eps, float(score)))
+
+
 def _hf_web_fallback_url(base_url: str) -> str:
     parsed = urlparse(base_url)
     host = parsed.netloc.lower()
@@ -243,7 +250,7 @@ async def run_single_task(env: AntiAtroposEnv, client: AsyncOpenAI, task_id: str
         log_step(step=step, action=action_str, reward=reward, done=bool(result.done), error=error)
 
     grade = grader.score()
-    score = max(0.0, min(1.0, float(grade.composite)))
+    score = _strict_score(float(grade.composite))
     success = score >= SUCCESS_SCORE_THRESHOLD
     return {
         "task_id": task_id,
@@ -256,28 +263,36 @@ async def run_single_task(env: AntiAtroposEnv, client: AsyncOpenAI, task_id: str
 
 async def run_all_tasks() -> None:
     _seed_everything(SEED)
-    task_id = TASK_NAME if TASK_NAME in {"task-1", "task-2", "task-3"} else "task-1"
+    all_tasks = ["task-1", "task-2", "task-3"]
+    run_single = os.getenv("ANTIATROPOS_RUN_SINGLE_TASK", "false").lower() == "true"
+    task_id = TASK_NAME if TASK_NAME in set(all_tasks) else "task-1"
+    tasks_to_run = [task_id] if run_single else all_tasks
     if not API_KEY:
-        raise RuntimeError("Missing API key (HF_TOKEN/API_KEY).")
+        raise RuntimeError("Missing API key (API_KEY/HF_TOKEN/OPENAI_API_KEY).")
 
     client = AsyncOpenAI(base_url=API_BASE_URL, api_key=API_KEY)
-    success = False
-    steps = 0
-    score = 0.0
-    rewards: List[float] = []
-
-    log_start(task=task_id, env=BENCHMARK, model=MODEL_NAME)
 
     try:
         async with open_env_with_ws_fallback(ENV_URL, MESSAGE_TIMEOUT_S) as env:
-            report = await run_single_task(env=env, client=client, task_id=task_id)
-            success = bool(report["success"])
-            steps = int(report["steps"])
-            score = float(report["score"])
-            rewards = list(report["rewards"])
+            for task in tasks_to_run:
+                success = False
+                steps = 0
+                score = 0.001
+                rewards: List[float] = []
+                log_start(task=task, env=BENCHMARK, model=MODEL_NAME)
+                try:
+                    report = await run_single_task(env=env, client=client, task_id=task)
+                    success = bool(report["success"])
+                    steps = int(report["steps"])
+                    score = _strict_score(float(report["score"]))
+                    rewards = list(report["rewards"])
+                except Exception as exc:
+                    print(f"[DEBUG] task={task} error={exc}", flush=True)
+                    score = 0.001
+                finally:
+                    log_end(task=task, success=success, steps=steps, score=score, rewards=rewards)
     finally:
         await client.close()
-        log_end(success=success, steps=steps, score=score, rewards=rewards)
 
 
 def main() -> None:
@@ -285,4 +300,4 @@ def main() -> None:
 
 
 if __name__ == "__main__":
-    main()
+    main()