> For the complete documentation index, see [llms.txt](https://guide.petoi.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://guide.petoi.com/extensible-modules/petoi-ai-vision-module/custom-petoi-vision-model/steps-on-how-to-get-started/step-1-choose-or-make-your-own-dataset/choose-a-pre-labeled-dataset.md).

# Choose a pre-labeled dataset

Don't want to spend hours drawing boxes around images? You don't have to!

For this path, we will use an open-source dataset. This means another developer or researcher has already done the hard work of collecting photos and labeling the objects inside them. Your job is simply to find the right dataset for your project and download it.

{% hint style="success" %}
**Goal:** Download a pre-labeled dataset that matches the objects your AI must detect.
{% endhint %}

***

## 🔍 Top Places to Find Datasets

There are massive online communities dedicated to sharing machine learning data. Here are the two best places to start your search:

* **Roboflow Universe:** This is the ultimate library specifically for computer vision. It has over 750,000 datasets that are already formatted for object detection. You can find almost anything here, from "playing cards" to "wild animals" to "traffic lights."
* **Kaggle:** A massive hub for data scientists. If you search the "Datasets" tab for "Object Detection," you will find thousands of high-quality, free collections.

{% hint style="info" %}
If you used Roboflow: Run the Roboflow Dataset section. The code will organize your files and automatically generate a `data.yaml` file. This file acts as a map, telling the AI exactly where your training images and labels are located.

If you used Kaggle: Run the Kaggle Dataset section. Because Kaggle datasets can sometimes have unpredictable formatting, these code cells will first validate that your dataset is in the correct YOLO format before organizing it for training.
{% endhint %}

***

## ⚠️ 3 Things to Check Before You Download

Not all datasets are created equal! When you find a dataset you like, check for these three things before you hit download:

1. The Export Format: Your dataset needs to speak the same language as your AI model. For most modern object detection models (like YOLO), you need to make sure you export or download the dataset in YOLO format.
2. Dataset Size: How many images are there? If a dataset only has 20 pictures, your AI will not have enough examples to learn from. Try to find a dataset that has at least a few hundred images.
3. The Classes: A "class" is the name of the object the AI is looking for. Check the dataset details to make sure it includes the exact classes you want to detect (e.g., if you want to detect fruit, make sure the classes are labeled "apple," "banana," etc.).

{% hint style="info" %}
Choose **YOLO** when the download site asks for an export format. Other formats may require conversion before training.
{% endhint %}

***

## 🛠️ Your Action Steps

{% stepper %}
{% step %}
Go to Roboflow Universe or Kaggle.
{% endstep %}

{% step %}
Search for an object you want your AI to recognize.
{% endstep %}

{% step %}
Verify that the dataset has enough images and the correct labels.
{% endstep %}

{% step %}
Download the dataset to your computer as a `.zip` file (make sure to select YOLO format if asked!).
{% endstep %}
{% endstepper %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://guide.petoi.com/extensible-modules/petoi-ai-vision-module/custom-petoi-vision-model/steps-on-how-to-get-started/step-1-choose-or-make-your-own-dataset/choose-a-pre-labeled-dataset.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.