Update handler.py

handler.py

@@ -39,9 +39,16 @@ class EndpointHandler:
 
         inputs = self.processor(text=given_text, return_tensors="pt")
 
+        run_time_processor = time.time() - start_time
+
+        start_time_speech = time.time()
+
         speech = self.model.generate_speech(inputs["input_ids"], self.speaker_embeddings, vocoder=self.vocoder)
+
+        run_time_speech = time.time() - start_time_speech
+
+        run_time_total = time.time() - start_time
         
-        run_time = time.time() - start_time
 
         # Return the expected response format
         return {
@@ -49,7 +56,9 @@ class EndpointHandler:
             "body": {
                 "audio": speech.numpy(),  # Consider encoding this to a suitable format
                 "sampling_rate": 16000,
-                "time_to_generate": str(run_time)
+                "run_time_processor": str(run_time_processor),
+                "run_time_speech": str(run_time_processor),
+                "run_time_total": str(run_time_processor),
             }
         }