Update functions.py

functions.py

@@ -12,39 +12,35 @@ def cargar_modalidades_tareas():
 
 # Función para actualizar el JSON desde HuggingFace y devolver tareas
 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 = ["text-classification", "image-classification", "speech-recognition"]
+    MODALIDAD_TAREAS = cargar_modalidades_tareas()  # Cargar JSON existente
     
-    modalidades_tareas = {}
     for task in MODALIDADES:
-        url = f"https://huggingface.co/api/datasets?task={task}&limit=5"
-        response = requests.get(url)
-        datasets = response.json()
+        url = f"https://huggingface.co/api/datasets?task={task}&full=true&limit=10"
+        response = requests.get(url).json()
         
-        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
-        
-        modalidad = "NLP" if task in ["text-classification", "token-classification", "translation", "summarization"] else \
-                   "Audio" if task in ["speech-recognition", "audio-classification", "music-classification"] else \
-                   "Vision" if task in ["image-classification", "object-detection", "image-segmentation"] else \
-                   "RAG"
-        
-        if modalidad not in modalidades_tareas:
-            modalidades_tareas[modalidad] = {"tareas": {}, "columnas_generales": []}
-        modalidades_tareas[modalidad]["tareas"][task] = list(columnas)
+        for dataset in response:
+            dataset_id = dataset["id"]
+            dataset_info = requests.get(f"https://huggingface.co/api/datasets/{dataset_id}").json()
+            
+            # Añadir datasets y columnas al JSON
+            if task not in MODALIDAD_TAREAS["NLP"]["tareas"]:
+                MODALIDAD_TAREAS["NLP"]["tareas"][task] = {
+                    "nombre": "Clasificación de texto",
+                    "columnas": ["text", "label"],
+                    "datasets": {}
+                }
+            
+            MODALIDAD_TAREAS["NLP"]["tareas"][task]["datasets"][dataset_id] = {
+                "columnas": list(dataset_info.get("features", {}).keys()),
+                "licencia": dataset.get("license", "unknown")
+            }
     
+    # Guardar JSON con indentación para evitar errores <button class="citation-flag" data-index="1">
     with open("modalidades_tareas.json", "w") as file:
-        json.dump(modalidades_tareas, file, indent=4)
+        json.dump(MODALIDAD_TAREAS, file, indent=4)
     
-    # Devolver lista de tareas actualizada
-    return [t for datos in modalidades_tareas.values() for t in datos["tareas"].keys()]
+    return "Metadatos actualizados: columnas y datasets almacenados."
 
     
 # Cargar modalidades y tareas