Datasetpublic

Kamran-56/prompt-refinement-dataset

Prompt Refinement Dataset Dataset Summary The Prompt Refinement Dataset is a curated collection of 4,349 input-output pairs designed to train language models to transform basic, vague prompts into high-quality, detailed, and structured prompts that elicit significantly better responses from AI systems. Each pair consists of a raw user-written prompt as the input and an expertly engineered version of the same prompt as the output — preserving the original intent… See the full description on the dataset page: https://huggingface.co/datasets/Kamran-56/prompt-refinement-dataset.

sourceHugging Facemitupdated 7mo agoView on Hugging Face

2likes16downloads

Dataset Card

Prompt Refinement Dataset

Dataset Summary

The Prompt Refinement Dataset is a curated collection of 4,349 input-output pairs designed to train language models to transform basic, vague prompts into high-quality, detailed, and structured prompts that elicit significantly better responses from AI systems.

Each pair consists of a raw user-written prompt as the input and an expertly engineered version of the same prompt as the output — preserving the original intent while adding role, context, format, tone, and constraints.

Dataset Structure

Columns

Column	Type	Description
`input`	string	Original raw prompt written by a user
`output`	string	Enhanced high-quality version of the same prompt

Example Row

Field	Value
input	`write a cover letter`
output	`Act as a professional career coach with 10 years of experience. Write a compelling cover letter for a software engineer applying to a FAANG company. The tone should be confident yet humble. Include a strong opening hook, 2 paragraphs highlighting technical skills and measurable impact, and a closing call to action. Keep it under 400 words.`

Dataset Creation

Source

Raw prompts were collected from a publicly available prompt rating dataset where prompts were rated on a scale of 1 to 5 based on quality.

Process

1. Collected prompts from a rated prompt dataset
2. Filtered and cleaned the raw prompts
3. Fed each prompt to a large language model (Llama 3.1 via Groq API)
   with a carefully designed prompt engineering instruction
4. Generated a high-quality enhanced version for each input prompt
5. Saved as input → output pairs for supervised fine-tuning

Instruction Used to Generate Outputs

You are an expert prompt engineer.
Transform the given basic prompt into a high quality, detailed,
and effective prompt that gets the best possible AI response.

Rules:
- Keep the original intent and topic
- Add role, context, format, tone, and constraints
- Do NOT answer the prompt, only improve it
- Return ONLY the improved prompt, nothing else

Intended Use

Primary Use

Fine-tuning small to medium language models to automatically enhance user-written prompts before sending them to AI systems.

Downstream Applications

—Chrome extensions that auto-enhance prompts on AI platforms
—API middleware that improves prompts before forwarding to LLMs
—Productivity tools that help non-technical users write better prompts

Out of Scope

This dataset is not intended for:

—Training models to answer questions
—General instruction following tasks
—Any use case outside of prompt quality improvement

Statistics

Property	Value
Total Pairs	1561
Language	English
Avg Input Length	~10-15 words
Avg Output Length	~50-100 words
Split	Train only

Limitations

—All prompts are in English only
—Enhanced outputs were generated by an LLM and may occasionally contain imperfect enhancements
—Dataset covers general purpose prompts — may not perform well on highly technical or domain-specific prompts

License

This dataset is released under the MIT License. You are free to use, modify, and distribute it with attribution.

Citation

If you use this dataset in your research or project, please cite:

bibtex

@dataset{prompt_refinement_dataset,
  author    = {Kamran},
  title     = {Prompt Refinement Dataset},
  year      = {2025},
  publisher = {HuggingFace},
  url       = {https://huggingface.co/datasets/Kamran-56/prompt-refinement-dataset}
}

Author

Created by Kamran as part of a prompt enhancement Chrome extension project that automatically improves user prompts on AI platforms like ChatGPT, Claude, and Gemini.