Update app.py

app.py

@@ -32,7 +32,7 @@ class TextChunker:
 
         sentences = sent_tokenize(text)
         if len(sentences) <= 1:
-            return [text] if test.strip() else []
+            return [text] if text.strip() else []
 
         contextualized = self._add_context(sentences, context_window)
         embeddings = self.model.encode(contextualized,batch_size=32,show_progress_bar=False)
@@ -63,8 +63,8 @@ class TextChunker:
 
 
     def _estimate_tokens(self, text: str) -> int:
-    # Fast approximation: 1 token ≈ 4 chars for English legal text
-    return len(text) // 4
+        # Fast approximation: 1 token ≈ 4 chars for English legal text
+        return len(text) // 4
 
     def _split_oversized(self, chunk: str, max_tokens: int, sent_tokenize) -> List[str]:
         """Split a chunk that exceeds max_tokens at sentence boundaries."""
@@ -208,7 +208,7 @@ async def lifespan(app: FastAPI):
 
 
     
-    models.update(loaded)
+    
     print("All models ready!")
     yield
     models.clear()
@@ -302,9 +302,7 @@ class BatchNLIResult(BaseModel):
 # ---------- Endpoints ----------
 @app.get("/health")
 def health():
-    return {"status": "ok",
-            "models_loaded": list(models.keys()),
-    }
+    return {"status": "ok",}
 
 
 @app.get("/memory")