| import torch | |
| from model import LSTMModel | |
| import gradio as gr | |
| import torch | |
| model = torch.load("LSTM_AdamW.pt") | |
| def classify_sentence(sentence): | |
| preds = model(sentence, pro = True, batched = False) | |
| dict = {"Negative": float(preds[0]), | |
| "Neutral": float(preds[1]), | |
| "Positive": float(preds[2])} | |
| return dict | |
| demo = gr.Interface(fn=classify_sentence, | |
| inputs=gr.Textbox(placeholder="Enter a sentence you'd like to classify here..."), | |
| outputs="label", | |
| examples=[["This is wonderful!"], ["It's okayish."], ["The product is not bad."], ["Unacceptable quality!"]]) | |
| demo.launch() |