(1) Abra el aviso de anaconda para crear un entorno:
conda create -n senta python=3.7
conda activate senta
(2) Instale los componentes:
pip install paddlepaddle==2.3.0 -i https://mirror.baidu.com/pypi/simple
pip install paddlehub==2.3.0 -i https://mirror.baidu.com/pypi/simple
hub install senta_bilstm==1.2.1
(3) Código de prueba:
import paddlehub as hub
senta = hub.Module(name="senta_bilstm")
input_dict = {
"text": ['这家餐厅很好吃', '这部电影真的很差劲']}
results = senta.sentiment_classify(data=input_dict)
for result in results:
print(result)
(4) Obtener el resultado:
{
'text': '这家餐厅很好吃', 'sentiment_label': 1, 'sentiment_key': 'positive', 'positive_probs': 0.9407, 'negative_probs': 0.0593}
{
'text': '这部电影真的很差劲', 'sentiment_label': 0, 'sentiment_key': 'negative', 'positive_probs': 0.02, 'negative_probs': 0.98}