{
  "default": {
    "description": "GlyphMatics complete synthetic canonical training dataset.",
    "splits": {
      "train": {
        "name": "train",
        "num_examples": 752
      },
      "validation": {
        "name": "validation",
        "num_examples": 83
      },
      "test": {
        "name": "test",
        "num_examples": 83
      }
    }
  }
}