metadata

task_categories:
  - summarization

AutoTrain Dataset for project: flant5finetune

Dataset Description

This dataset has been automatically processed by AutoTrain for project flant5finetune.

Languages

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

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "text": "Describe this image of a serene lake and turn it into a starry night scene.",
    "target": "Tools: Image captioning, Image editing\nInput: {Image}, turn it into a starry night scene\n"
  },
  {
    "text": "Generate a melodic rock track for an energetic outdoor adventure video.",
    "target": "Tools: Music generation\nInput: melodic rock track for an energetic outdoor adventure video\n"
  }
]

Dataset Fields

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

{
  "text": "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	23
valid	6