Actualización de modelo

config.json

@@ -1,20 +1,10 @@
 {
-  "architectures": ["AutoModelForSequenceClassification", "AutoModelForCausalLM"],
-  "model_type": "bert",
-  "classification_model_name": "delineiro/soflexpoc",
-  "generation_model_name": "delineiro/soflexpoc",
+  "model_name": "Spanish Text Classification Model",
+  "model_type": "sklearn",
+  "framework": "sklearn",
+  "task": "text-classification",
   "language": "es",
-  "max_position_embeddings": 512,
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "pad_token_id": 0,
-  "vocab_size": 30522,
-  "hidden_size": 768,
-  "intermediate_size": 3072,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "attention_probs_dropout_prob": 0.1,
-  "max_length": 100,
-  "do_sample": true,
-  "num_return_sequences": 1
-}
+  "vectorizer": "TfidfVectorizer",
+  "classifier": "MultinomialNB",
+  "version": "1.0.0"
+}

inference.py

@@ -1,37 +1,20 @@
-import joblib
-import numpy as np
-
-# Cargar el vectorizador
-try:
-    with open('vectorizer.pkl', 'rb') as f:
-        vectorizer = joblib.load(f)
-except Exception as e:
-    print(f"Error al cargar el vectorizador: {e}")
 
-# Cargar el modelo
-try:
-    with open('model.pkl', 'rb') as f:
-        model = joblib.load(f)
-except Exception as e:
-    print(f"Error al cargar el modelo: {e}")
+import joblib
 
-def preprocess(text):
-    """Preprocesa el texto para la inferencia."""
-    try:
-        return vectorizer.transform([text])
-    except Exception as e:
-        print(f"Error en el preprocesamiento: {e}")
+def load_model_and_vectorizer():
+    model = joblib.load('model.pkl')
+    vectorizer = joblib.load('vectorizer.pkl')
+    return model, vectorizer
 
 def predict(text):
-    """Realiza la predicción a partir del texto ingresado."""
-    try:
-        X = preprocess(text)
-        return model.predict(X)
-    except Exception as e:
-        print(f"Error en la predicción: {e}")
+    model, vectorizer = load_model_and_vectorizer()
+    text_vectorized = vectorizer.transform([text])
+    prediction = model.predict(text_vectorized)
+    return prediction[0]
 
-if __name__ == "__main__":
-    # Prueba del modelo
-    test_text = "Texto de prueba para predecir"
-    result = predict(test_text)
-    print(f"Predicción: {result}")
+if __name__ == '__main__':
+    # Example usage
+    text = "Ejemplo de declaraci�n"
+    result = predict(text)
+    print(f"Categor�a predicha: {result}")
+    

model.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6566d630fd91a6406a0b60465fd1941d1a475655c53088aeae71e38c4662515d
+oid sha256:e5910e6d9fdc5fd8f6ed1cceeda5bfb8dac91905e61c0bd01dbf7a6c39b416f6
 size 7103071

vectorizer.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e74c46f64f2513cc152fc1751b0fa1d8cb0015576b800a7ec4d3ded27b490bb6
+oid sha256:9d38d53585deb8d0789ce2a229e8a000849900a1cedf8c91b9c9d7d1d2f76cfb
 size 1123551