Update app.py

app.py

@@ -7,6 +7,7 @@ import os
 from dotenv import load_dotenv
 from pydantic import BaseModel
 import requests
+import traceback
 
 load_dotenv()
 HUGGINGFACE_TOKEN = os.getenv("HUGGINGFACE_TOKEN")
@@ -41,6 +42,7 @@ def load_model(model_config):
             return model
         except Exception as e:
             print(f"Error loading model {model_name}: {e}")
+            traceback.print_exc()
             models[model_name] = None
             return None
 
@@ -72,21 +74,19 @@ def generate_model_response(model, inputs, max_tokens_per_part):
             return []
         full_response = ""
         responses = []
-        tokens_generated = 0
-        while True:
-            response_part = model(inputs, max_tokens=max_tokens_per_part, stop=["\n\n"])
-            text = response_part['choices'][0]['text']
-            if not text.strip():
-                break
-            full_response += text
-            tokens_generated += len(response_part['choices'][0]['token'])
-            responses.append(remove_duplicates(text))
-            if "eos_token" in response_part['choices'][0]['token']:
-                break
-            inputs = ""
+        response = model(inputs, max_tokens=max_tokens_per_part, stop=["\n\n"])
+        if 'choices' not in response or len(response['choices']) == 0 or 'text' not in response['choices'][0]:
+            return [f"Error: Invalid model response format"]
+
+        text = response['choices'][0]['text']
+        if text:
+          responses.append(remove_duplicates(text))
+
         return responses
+
     except Exception as e:
         print(f"Error generating response: {e}")
+        traceback.print_exc()
         return [f"Error: {e}"]
 
 app = FastAPI()