import gradio as gr
import numpy as np
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

id2label = {0: 'anger', 1: 'anticipation', 2: 'disgust', 3: 'fear', 4: 'joy', 5: 'love', 6: 'optimism', 7: 'pessimism', 8: 'sadness', 9: 'surprise', 10: 'trust'}
tokenizer = AutoTokenizer.from_pretrained("winain7788/bert-finetuned-sem_eval-english")
model = AutoModelForSequenceClassification.from_pretrained("winain7788/bert-finetuned-sem_eval-english")

async def get_sentiment(text):
    encoding = tokenizer(text, return_tensors="pt")
    encoding = {k: v.to(model.device) for k,v in encoding.items()}

    outputs = model(**encoding)
    logits = outputs.logits
    logits.shape
    # apply sigmoid + threshold
    sigmoid = torch.nn.Sigmoid()
    probs = sigmoid(logits.squeeze().cpu())
    predictions = np.zeros(probs.shape)
    predictions[np.where(probs >= 0.5)] = 1

    # turn predicted id's into actual label names
    predicted_labels = [id2label[idx] for idx, label in enumerate(predictions) if label == 1.0]
    return predicted_labels

demo = gr.Interface(fn=get_sentiment, inputs="text", outputs="json")

demo.launch()