File size: 789 Bytes
a72b0a2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
# imports
from transformers import pipeline
import gradio as gr

# define nlp mask
model = "siebert/sentiment-roberta-large-english"
nlp = pipeline(model=model)  # set device=0 to use GPU (CPU default, -1)

# Inference
def inference(sentence):
  preds = nlp(sentence)
  pred_sentiment = preds[0]["label"]
  pred_score = preds[0]["score"]
  return pred_sentiment, pred_score
  
# launch app
gr.Interface(inference,
             inputs=[gr.inputs.Textbox(label="Sentiment to predict", default="I love this!")],
             outputs=[gr.outputs.Textbox(type="auto", label="Predicted sentiment"),
                      gr.outputs.Textbox(type="auto", label="Predicted score")],
             description="Sentiment analysis",
             allow_flagging=False,
             ).launch(debug=True)