Dataset Overview

Dataset Information

• Features:

  • sentence: type string, representing the sentence.
  • label: type categorical, containing two classes:
    • 0: negative sentiment.
    • 1: positive sentiment.
  • idx: type integer, representing the index.

• Dataset Split:

  • train: training set, 68,221 samples, occupying 4,787,855 bytes.
  • validation: validation set, 872 samples, occupying 106,252 bytes.
  • test: test set, 1,821 samples, occupying 216,640 bytes.

• Dataset Size:

  • Download size: 3,403,184 bytes.
  • Total dataset size: 5,110,747 bytes.

Configuration

• Configuration name: default
• Data files:
  • train: path data/train-*
  • validation: path data/validation-*
  • test: path data/test-*