# Train instruct pix2pix task with dreambooth

**URL:** <https://discuss.huggingface.co/t/train-instruct-pix2pix-task-with-dreambooth/49675>\
**Category:** Intermediate\
**Created:** [August 6, 2023, 12:12pm UTC](https://discuss.huggingface.co/t/train-instruct-pix2pix-task-with-dreambooth/49675 "2023-08-06T12:12:36Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![mzeynali](https://avatars.discourse-cdn.com/v4/letter/m/d07c76/32.png) [@mzeynali](https://discuss.huggingface.co/u/mzeynali)\
**Post date:** [August 6, 2023, 12:12pm UTC](https://discuss.huggingface.co/t/train-instruct-pix2pix-task-with-dreambooth/49675/1 "2023-08-06T12:12:36Z")

</div>

Hi guys,  
I have a dataset has multi-category, and each of the categories have a few examples, and the categories are relatively similar. Is it possible to train dreambooth-lora for this? I want to know the dreambooth-lora can be reach accuracy to instruct-pix2pix method?  
My task is add objects to images.