Update functions.py

functions.py

@@ -1,9 +1,50 @@
 # functions.py
 import json
-import requests  # Añadido <button class="citation-flag" data-index="1">
-import matplotlib.pyplot as plt  # Añadido <button class="citation-flag" data-index="1">
-from datasets import load_dataset
+import requests
 
+def actualizar_modalidades_tareas_desde_huggingface():
+    MODALIDADES = ["text-classification", "token-classification", "translation", "summarization",
+                   "speech-recognition", "audio-classification", "music-classification",
+                   "image-classification", "object-detection", "image-segmentation",
+                   "question-answering", "information-retrieval"]
+    
+    modalidades_tareas = {}
+    for task in MODALIDADES:
+        url = f"https://huggingface.co/api/datasets?task={task}&limit=5"  # Buscar datasets por tarea <button class="citation-flag" data-index="1">
+        response = requests.get(url)
+        datasets = response.json()
+        
+        # Extraer columnas comunes de los datasets
+        columnas = set()
+        for dataset in datasets:
+            try:
+                info = requests.get(f"https://huggingface.co/api/datasets/{dataset['id']}").json()
+                columnas.update(info["features"].keys())
+            except:
+                continue
+        
+        # Asignar a modalidades
+        if task in ["text-classification", "token-classification", "translation", "summarization"]:
+            modalidad = "NLP"
+        elif task in ["speech-recognition", "audio-classification", "music-classification"]:
+            modalidad = "Audio"
+        elif task in ["image-classification", "object-detection", "image-segmentation"]:
+            modalidad = "Vision"
+        elif task in ["question-answering", "information-retrieval"]:
+            modalidad = "RAG"
+        
+        # Añadir al diccionario
+        if modalidad not in modalidades_tareas:
+            modalidades_tareas[modalidad] = {"tareas": {}, "columnas_generales": []}
+        modalidades_tareas[modalidad]["tareas"][task] = list(columnas)
+    
+    # Guardar JSON
+    with open("modalidades_tareas.json", "w") as file:
+        json.dump(modalidades_tareas, file, indent=4)
+    
+    return "Archivo actualizado con tareas y columnas de HuggingFace."
+
+    
 # Cargar modalidades y tareas
 def cargar_modalidades_tareas():
     with open("modalidades_tareas.json", "r") as file: