Update the lifecycle to load the model only from local files.

gliner_inference_server/main.py

@@ -24,7 +24,10 @@ async def lifespan(app: FastAPI):
     print("Loading GLiNER model...")
     model_name = os.getenv("MODEL_NAME", "knowledgator/gliner-multitask-large-v0.5")
     device = "cuda" if torch.cuda.is_available() else "cpu"
-    model = GLiNER.from_pretrained(model_name).to(device)
+    model = GLiNER.from_pretrained(
+        model_name,
+        local_files_only = True
+    ).to(device)
     print(f"Model loaded on {device}")
     yield
     print("Shutting down...")