Hello!
As I’ve been annotating data it recommends I split the data into the proper test, train, valid sets. However, I’d like to re-randomize / re-distribute my dataset into those categories. Is there an easy way to do that with roboflow.
Yes, Roboflow allows you to easily re-distribute your dataset into training, validation, and testing sets. Here’s how you can do it:
- When generating a new version of your dataset, click on “Step 2” (Train/Test Split).
- Next, select “Rebalance”.
- Adjust the sliders to select your new desired Train/Validation/Test split.
- Select “Save” and continue on with the Preprocessing and Augmentation steps to generate a version of your dataset with the new train/validation/test split.
Please note that the Preprocessing step of “Tiling” will increase the number of images in your dataset. The Augmentation steps will also generate more examples of your images. Application of Tiling, or augmentations, will result in more total images in your training set. To keep your train/valid/test split balance exactly the same as the one you selected during “Rebalancing,” do not apply Tiling (preprocessing) or any augmentation steps.
You can also choose to upload images to only training, validation, or test set if desired.
For more details, you can refer to this help page.
Additionally, Roboflow has also introduced in-app changes that allow you to manage your train test splits as you are working through the computer vision workflow. You can find more information about this in this blog post.
Hi! I’m Roboflow’s AI powered bot. This is an automated response powered using AI with knowledge from our docs, blog, knowledge base, and previous forum responses. If you find this to be a helpful answer, please mark it as the solution using the checkbox icon below. If this doesn’t solve your issue, please follow up with what you’ve tried/why, and the community will continue to chime in as usual.