Update app.py

app.py

@@ -12,7 +12,8 @@ import psutil
 import asyncio
 import torch
 from tenacity import retry, stop_after_attempt, wait_fixed
-from huggingface_hub import HfApi, RepositoryNotFoundError
+from huggingface_hub import HfApi
+from huggingface_hub.utils import RepositoryNotFoundError
 from dotenv import load_dotenv
 
 # Cargar variables de entorno
@@ -40,10 +41,10 @@ pipeline_instance = None  # Solo un pipeline
 initialization_complete = False
 
 def initialize_model():
-    global pipeline_instance
+    global pipeline_instance, initialization_complete
     try:
-        logger.info("Initializing the base model and tokenizer.")
-        base_model_repo = "meta-llama/Llama-3.2-1B"
+        logger.info("Initializing the GPT-2 model and tokenizer.")
+        base_model_repo = "gpt2"  # Puedes elegir variantes como "gpt2-medium", "gpt2-large", etc.
         model = AutoModelForCausalLM.from_pretrained(
             base_model_repo,
             cache_dir=CACHE_DIR,
@@ -58,7 +59,8 @@ def initialize_model():
             tokenizer=tokenizer,
             device=0 if torch.cuda.is_available() else -1
         )
-        logger.info("Model and tokenizer initialized successfully.")
+        logger.info("GPT-2 model and tokenizer initialized successfully.")
+        initialization_complete = True
     except Exception as e:
         logger.error(f"Error initializing model and tokenizer: {e}", exc_info=True)
         sys.exit(1)
@@ -77,7 +79,7 @@ def download_dataset(dataset_name):
 def upload_model_to_hub():
     try:
         api = HfApi()
-        model_repo = "Yhhxhfh/test"
+        model_repo = "Yhhxhfh/Hhggg"  # Reemplaza con tu repositorio en Hugging Face Hub
         try:
             api.repo_info(repo_id=model_repo)
             logger.info(f"Model repository {model_repo} already exists.")
@@ -95,11 +97,11 @@ def create_example_usage(dataset_name):
     try:
         logger.info(f"Creating example usage for dataset {dataset_name}")
         example_prompts = [
-            "Translate the following catering menu from English to French:",
-            "Generate a catering menu for a wedding with vegetarian options:",
-            "Convert the following catering menu to a gluten-free version:",
-            "Provide a detailed catering menu for a corporate event including desserts:",
-            "Generate a children's birthday party catering menu with allergen-free items:"
+            "Once upon a time,",
+            "In a world where AI rules,",
+            "The future of technology is",
+            "Explain the concept of",
+            "Describe a scenario where"
         ]
         examples = []
         for prompt in example_prompts:
@@ -139,22 +141,16 @@ logger.info(f"Using up to {max_concurrent_downloads} concurrent workers for down
 executor = concurrent.futures.ThreadPoolExecutor(max_workers=max_concurrent_downloads)
 
 async def download_and_process_datasets():
-    global initialization_complete
-    try:
-        dataset_names = list_datasets()
-        logger.info(f"Found {len(dataset_names)} datasets to download.")
-        loop = asyncio.get_event_loop()
-        tasks = []
-        for dataset_name in dataset_names:
-            task = loop.run_in_executor(executor, download_dataset, dataset_name)
-            tasks.append(task)
-        await asyncio.gather(*tasks)
-        unify_datasets()
-        upload_model_to_hub()
-        initialization_complete = True
-        logger.info("All initialization tasks completed successfully.")
-    except Exception as e:
-        logger.error(f"Error during dataset processing: {e}", exc_info=True)
+    dataset_names = list_datasets()
+    logger.info(f"Found {len(dataset_names)} datasets to download.")
+    loop = asyncio.get_event_loop()
+    tasks = []
+    for dataset_name in dataset_names:
+        task = loop.run_in_executor(executor, download_dataset, dataset_name)
+        tasks.append(task)
+    await asyncio.gather(*tasks)
+    unify_datasets()
+    upload_model_to_hub()
 
 # Inicializar FastAPI
 app = FastAPI()
@@ -162,7 +158,7 @@ app = FastAPI()
 # Configuración de CORS
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],  # Cambia esto según tus necesidades
+    allow_origins=["*"],  # Para mayor seguridad, especifica los orígenes permitidos
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"]
@@ -174,17 +170,17 @@ message_history = []
 async def startup_event():
     logger.info("Application startup initiated.")
     loop = asyncio.get_event_loop()
-    # Crear una tarea en segundo plano para inicializar el modelo y descargar datasets
-    loop.create_task(run_initialization(loop))
-    logger.info("Background initialization tasks started.")
+    # Iniciar tareas en segundo plano sin bloquear el servidor
+    asyncio.create_task(run_initialization(loop))
+    logger.info("Startup tasks initiated.")
 
 async def run_initialization(loop):
-    global initialization_complete
     try:
         # Inicializar el modelo en un hilo separado
         await loop.run_in_executor(None, initialize_model)
         # Descargar y procesar datasets
         await download_and_process_datasets()
+        logger.info("All startup tasks completed successfully.")
     except Exception as e:
         logger.error(f"Error during startup tasks: {e}", exc_info=True)
 
@@ -197,7 +193,7 @@ async def index():
         <!-- Existing head content -->
         <meta charset="UTF-8">
         <meta name="viewport" content="width=device-width, initial-scale=1.0">
-        <title>ChatGPT Chatbot</title>
+        <title>GPT-2 Chatbot</title>
         <style>
             /* Existing styles */
             /* Add styles for the model selector */
@@ -290,7 +286,7 @@ async def index():
     </head>
     <body>
         <div class="container">
-            <h1 style="text-align: center;">ChatGPT Chatbot</h1>
+            <h1 style="text-align: center;">GPT-2 Chatbot</h1>
             <div class="chat-container" id="chat-container">
                 <div class="chat-box" id="chat-box">
                 </div>
@@ -361,7 +357,8 @@ async def index():
                 })
                 .catch(error => {
                     console.error('Error:', error);
-                    appendMessage('bot', error.message || 'An error occurred. Please try again later.');
+                    saveMessage('bot', "Sorry, I'm not available right now. Please try again later.");
+                    appendMessage('bot', "Sorry, I'm not available right now. Please try again later.");
                 });
             }
 
@@ -393,8 +390,8 @@ async def autocomplete(q: str = Query(..., title='query')):
         return {"result": [response]}
     except Exception as e:
         logger.error(f"Ignored error in autocomplete: {e}", exc_info=True)
-        raise HTTPException(status_code=500, detail="An error occurred while processing your request.")
+        return {"result": []}
 
 if __name__ == '__main__':
-    port = int(os.getenv("PORT", 7860))
+    port = 7860  # Configurar FastAPI para que inicie en el puerto 7860
     uvicorn.run(app=app, host='0.0.0.0', port=port)