README.md

metadata

language:
  - en
  - zh
task_categories:
  - translation

AutoTrain Dataset for project: lojban-translation

Dataset Description

This dataset has been automatically processed by AutoTrain for project lojban-translation.

Languages

The BCP-47 code for the dataset's language is en2zh.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "source": "I read the poem for my child.",
    "target": "mi tcidu lo pemci te cu'u le panzi be mi"
  },
  {
    "source": "Jim is learning how to drive a car.",
    "target": "la jim cilre fi lo nu klasazri lo karce"
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "source": "Value(dtype='string', id=None)",
  "target": "Value(dtype='string', id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	8000
valid	2000