Update app.py

app.py

@@ -5,20 +5,20 @@ from concurrent.futures import ThreadPoolExecutor, as_completed
 import uvicorn
 import re
 from dotenv import load_dotenv
-from spaces import GPU
+import gradio as gr
 from spaces import ZeroGPU
 
 load_dotenv()
 
-# Initialize ZeroGPU
+app = FastAPI()
+
+# Inicializar ZeroGPU
 try:
     ZeroGPU.initialize()
 except Exception as e:
     print(f"ZeroGPU initialization failed: {e}")
 
-app = FastAPI()
-
-# Global data dictionary
+# Diccionario global para almacenar modelos y tokens
 global_data = {
     'models': {},
     'tokens': {
@@ -107,25 +107,17 @@ def remove_duplicates(text):
     seen_lines = set()
     for line in lines:
         if line not in seen_lines:
-            seen_lines.add(line)
             unique_lines.append(line)
+            seen_lines.add(line)
     return '\n'.join(unique_lines)
 
-def remove_repetitive_responses(responses):
-    seen = set()
-    unique_responses = []
-    for response in responses:
-        normalized_response = remove_duplicates(response['response'])
-        if normalized_response not in seen:
-            seen.add(normalized_response)
-            unique_responses.append({'model': response['model'], 'response': normalized_response})
-    return unique_responses
-
+@spaces.GPU(duration=0)
 def generate_model_response(model, inputs, top_k, top_p, temperature):
     try:
-        return model.generate(inputs, top_k=top_k, top_p=top_p, temperature=temperature)
+        response = model.generate(inputs, top_k=top_k, top_p=top_p, temperature=temperature)
+        return remove_duplicates(response)
     except Exception as e:
-        print(f"Error generating response with model {model}: {e}")
+        print(f"Error generating model response: {e}")
         return ""
 
 @app.post("/generate")
@@ -152,5 +144,12 @@ async def process_request(request: Request, call_next):
         print(f"Request error: {e}")
         raise HTTPException(status_code=500, detail="Internal Server Error")
 
+def remove_repetitive_responses(responses):
+    unique_responses = {}
+    for response in responses:
+        if response['model'] not in unique_responses:
+            unique_responses[response['model']] = response['response']
+    return unique_responses
+
 if __name__ == "__main__":
     uvicorn.run(app, host="0.0.0.0", port=8000)