File size: 876 Bytes
65f03a1 fae15ea 65f03a1 c9e14db 582c5da c9e14db 582c5da 65f03a1 49891f3 65f03a1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 |
import os
import gradio as gr
import torch
from transformers.models.bert import BertTokenizer, BertForSequenceClassification
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("sundea/text1")
model = AutoModelForSequenceClassification.from_pretrained("sundea/text1")
model.eval()
def get_output(text):
output=[]
model_input = tokenizer(text, return_tensors="pt", padding=True)
model_output = model(**model_input, return_dict=False)
prediction = torch.argmax(model_output[0].cpu(), dim=-1)
prediction = [p.item() for p in prediction]
for i in range(len(prediction)):
if prediction[i]==1:
output.append("骂人")
else:
output.append('非骂人')
return output
demo=gr.Interface(fn=get_output,inputs='text',outputs='text')
demo.launch() |