Model clearing

app.py

@@ -9,6 +9,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 from huggingface_hub import login
 from dotenv import load_dotenv
 import os
+import uvicorn
 
 import logging
 # Configure logging
@@ -65,8 +66,9 @@ async def lifespan(app: FastAPI):
         raise e
     # Release resources when the app is stopped
     yield
-    #model.clear()
-    #tokenizer.clear()
+    del model
+    del tokenizer
+    cuda.empty_cache()
 
 
 # Setup our app
@@ -130,6 +132,7 @@ async def predict(q: str):
         )
 
         # Get and decode the output
+        print(generated_ids)
         output_ids = generated_ids[0][-1]
         logger.debug(output_ids)
         #[len(model_inputs.input_ids[0]) :]