dipenbhuva/home-diy-repair-qa
Home DIY Repair Q&A A synthetic dataset of 5,000 Q&A pairs covering common home DIY repair scenarios. Each example includes a detailed step-by-step answer, required tools, safety warnings, and practical tips. Dataset Purpose This dataset is built for: Instruction fine-tuning — train language models to give detailed, safe, and actionable home repair guidance Retrieval-Augmented Generation (RAG) — build a knowledge base for home repair assistants Question answering… See the full description on the dataset page: https://huggingface.co/datasets/dipenbhuva/home-diy-repair-qa.
Home DIY Repair Q&A
A synthetic dataset of 5,000 Q&A pairs covering common home DIY repair scenarios. Each example includes a detailed step-by-step answer, required tools, safety warnings, and practical tips.
Dataset Purpose
This dataset is built for:
- Instruction fine-tuning — train language models to give detailed, safe, and actionable home repair guidance
- Retrieval-Augmented Generation (RAG) — build a knowledge base for home repair assistants
- Question answering — train or evaluate models on structured how-to responses
- Embedding fine-tuning — learn representations of repair problems and solutions
Dataset Structure
Each example is a structured Q&A pair about a home repair problem.
Splits
Categories
The dataset covers 5 categories, evenly distributed (1,000 each before splitting):
Usage
Loading the Dataset
from datasets import load_dataset
dataset = load_dataset("dipenbhuva/home-diy-repair-qa")
train = dataset["train"]
eval_set = dataset["eval"]
print(train[0])Instruction Fine-tuning
Format examples as instruction-response pairs:
def format_for_finetuning(example):
return {
"instruction": example["question"],
"response": example["answer"],
}
formatted = dataset["train"].map(format_for_finetuning)Using Structured Fields
Access individual components for custom formatting:
example = dataset["train"][0]
print("Problem:", example["equipment_problem"])
print("Tools:", example["tools_required"])
print("Safety:", example["safety_info"])
for i, step in enumerate(example["steps"], 1):
print(f" {i}. {step}")
print("Tips:", example["tips"])Example
{
"id": "qa_00001",
"category": "appliance_repair",
"question": "My washing machine is not filling with water. How do I fix this?",
"answer": "Before you begin, an important safety note: Turn off both water supply valves before disconnecting hoses...",
"equipment_problem": "Washing machine not filling with water",
"tools_required": ["adjustable pliers", "toothbrush", "bucket", "towels"],
"steps": [
"Check that both the hot and cold water supply faucets behind the machine are fully open",
"Inspect the inlet hoses for kinks or pinches that could restrict water flow",
"Turn off the water supply and disconnect the hoses from the back of the machine",
"Clean the inlet filter screens inside the hose connections with a toothbrush",
"Reconnect the hoses hand-tight plus a quarter turn with pliers",
"Turn the supply valves back on and run a fill cycle to verify water enters the machine"
],
"safety_info": "Turn off both water supply valves before disconnecting hoses. Have towels and a bucket ready.",
"tips": [
"Clogged inlet screens are the most common cause of fill problems — clean them every year",
"If the machine fills slowly, the water supply valves may not be fully open"
]
}Generation
This dataset was synthetically generated using OpenAI's API with 5 diverse prompt templates, followed by Pydantic schema validation, quality validation (6 rules), and LLM-as-Judge failure analysis. The pipeline achieved a 100% structural validity rate and 80% quality pass rate, with the primary failure mode being overcomplicated solutions.
Limitations
- Synthetic data: Generated by an LLM, not written by professional repair technicians. Instructions should be verified before use on real repairs.
- English only: All text is in English.
- General guidance: Answers provide general approaches and may not cover all makes/models or local building codes.
- Safety: While safety info is included, always consult a licensed professional for electrical, gas, or structural work.
Citation
@dataset{home_diy_repair_qa,
title={Home DIY Repair Q&A},
author={dipenbhuva},
year={2025},
url={https://huggingface.co/datasets/dipenbhuva/home-diy-repair-qa}
}