README.md

metadata

annotations_creators: []
language:
  - en
language_creators: []
license: []
multilinguality:
  - monolingual
pretty_name: adaption-alpaca_instructions
size_categories:
  - 10K<n<100K
source_datasets:
  - extended|https://huggingface.co/datasets/tatsu-lab/alpaca
tags:
  - adaption
  - instruction-tuning
  - writing-editing-communication
  - math
  - language
task_categories: []
task_ids: []

This dataset is a remastered version of this dataset prepared using Adaption's Adaptive Data platform.

adaption-alpaca_instructions

This dataset contains 52,000 instruction-following examples generated by OpenAI's text-davinci-003 engine using the Self-Instruct framework. Each sample consists of a prompt providing a specific task, such as writing, classification, or rewriting, paired with a high-quality completion. It is designed specifically for instruction-tuning language models to improve their ability to follow diverse user commands.

Dataset size

There are 22,155 data points in this dataset. This is an instruction tuning dataset.

Quality of Remastered Dataset

The final quality is B, with a relative quality improvement of 48.0%.

Domain

• Writing-editing-communication (20%)
• Math (10%)
• Language (8%)

Language

• English (100%)

Tone

• Informative (28%)
• Clear (14%)
• Creative (8%)

Evaluation Results

• Quality Gains:

  

• Grade Improvement:

  

• Percentile Chart: