twi_wordsim353

References:

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:twi_wordsim353')
  • Description:
A translation of the word pair similarity dataset wordsim-353 to Twi.

The dataset was presented in the paper
Alabi et al.: Massive vs. Curated Embeddings for Low-Resourced
Languages: the Case of Yorùbá and Twi (LREC 2020).
  • License: No known license
  • Version: 0.0.0
  • Splits:
Split Examples
'test' 274
  • Features:
{
    "twi1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "twi2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "similarity": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}