ab3llini

import torch
from utils.data.paths import resources_path
from torch.utils.data import random_split
from transformers.optimization import AdamW
from ignite.engine import Engine, Events
from ignite.metrics import RunningAverage, Accuracy, Precision
from ignite.handlers import ModelCheckpoint, EarlyStopping
from torch.utils.data import DataLoader
from ignite.contrib.handlers import ProgressBar
from transformers import BertForSequenceClassification, AutoTokenizer

ab3llini

import torch
from torch import nn
import os
from transformers import BertForSequenceClassification, AutoTokenizer


def extract_sentiment(model, tokenizer, text, device):
    # Encode the text, create a tensor and move to device.
    tensor = torch.tensor(tokenizer.encode(text, add_special_tokens=True)).unsqueeze(0).long().to(device)

ab3llini

import torch
from torch.utils.data import Dataset
import json


class TweetDataset(Dataset):
    def __init__(self, path, device):
        self.device = device

        # Load the JSON file containing our pre-processed data

ab3llini

from multiprocessing import cpu_count
import pandas as pd
import numpy as np
from multiprocessing import Pool
from transformers import AutoTokenizer
import spacy


class TweetProcessor: