Update handler.py

133bed4 verified about 2 years ago

1.81 kB

from typing import Dict, List, Any
from transformers import pipeline
import logging

# Configuración básica del logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

class EndpointHandler():
    def __init__(self, path=""):
        logger.info(f"Modelo cargado desde el path: {path}")
        # Cargar el pipeline con el modelo especificado
        try:
            self.pipeline = pipeline(model=path, truncation=True)
        except Exception as e:
            logger.exception(f"Error cargando el modelo desde el path {path}: {e}")
            path = "AndresR2909/finetuning-bert-text-classification"
            self.pipeline = pipeline(model=path, truncation=True)


    def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
        """
       data args:
            inputs (:obj: `str`)
            date (:obj: `str`)
      Return:
            A :obj:`list` | `dict`: will be serialized and returned
        """
        # get inputs
        input = data.get("inputs",data)
        date = data.get("date", None)

        # Textos para análisis
        texts = [input]
        # Realizar la inferencia
        outputs = self.pipeline(texts)
        
        # Ajustar las etiquetas a 0 y 1
        label_mapping = {"LABEL_0": 0, "LABEL_1": 1}
        label_names = {0: "sin_intencion", 1: "intencion_suicida"}
        
        # Modificar las salidas con el formato deseado
        adjusted_results = [
            {
                "input": text,
                "clasiffication": str(label_mapping[result['label']]),  # "clasiffication" como string
                "label": label_names[label_mapping[result['label']]]    # Mapeo de etiquetas a nombres
            }
            for result, text in zip(outputs, texts)
        ]
        return adjusted_results

	from typing import Dict, List, Any
	from transformers import pipeline
	import logging

	# Configuración básica del logging
	logging.basicConfig(level=logging.INFO)
	logger = logging.getLogger(__name__)

	class EndpointHandler():
	def __init__(self, path=""):
	logger.info(f"Modelo cargado desde el path: {path}")
	# Cargar el pipeline con el modelo especificado
	try:
	self.pipeline = pipeline(model=path, truncation=True)
	except Exception as e:
	logger.exception(f"Error cargando el modelo desde el path {path}: {e}")
	path = "AndresR2909/finetuning-bert-text-classification"
	self.pipeline = pipeline(model=path, truncation=True)


	def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
	"""
	data args:
	inputs (:obj: `str`)
	date (:obj: `str`)
	Return:
	A :obj:`list` \| `dict`: will be serialized and returned
	"""
	# get inputs
	input = data.get("inputs",data)
	date = data.get("date", None)

	# Textos para análisis
	texts = [input]
	# Realizar la inferencia
	outputs = self.pipeline(texts)

	# Ajustar las etiquetas a 0 y 1
	label_mapping = {"LABEL_0": 0, "LABEL_1": 1}
	label_names = {0: "sin_intencion", 1: "intencion_suicida"}

	# Modificar las salidas con el formato deseado
	adjusted_results = [
	{
	"input": text,
	"clasiffication": str(label_mapping[result['label']]), # "clasiffication" como string
	"label": label_names[label_mapping[result['label']]] # Mapeo de etiquetas a nombres
	}
	for result, text in zip(outputs, texts)
	]
	return adjusted_results