# How can I use my custom dataset on PYNQ Z2 for QNN, which is already trained on Pascal VOC dataset?

**URL:** https://discuss.pynq.io/t/how-can-i-use-my-custom-dataset-on-pynq-z2-for-qnn-which-is-already-trained-on-pascal-voc-dataset/4612
**Category:** Support
**Created:** [August 29, 2022, 8:02am UTC](https://discuss.pynq.io/t/how-can-i-use-my-custom-dataset-on-pynq-z2-for-qnn-which-is-already-trained-on-pascal-voc-dataset/4612 "2022-08-29T08:02:43Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![Vineeth\_C\_Johnson](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/vineeth_c_johnson/32/3546_2.png) [@Vineeth\_C\_Johnson](https://discuss.pynq.io/u/Vineeth_C_Johnson)
#### Post date: [August 29, 2022, 8:02am UTC](https://discuss.pynq.io/t/how-can-i-use-my-custom-dataset-on-pynq-z2-for-qnn-which-is-already-trained-on-pascal-voc-dataset/4612/1 "2022-08-29T08:02:43Z")

</div>

I want to use my custom dataset for object detection on PYNQ Z2 using QNN.

---

<div class="post-metadata">

### Author: ![briansune](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/briansune/32/3087_2.png) [@briansune](https://discuss.pynq.io/u/briansune)
#### Post date: [August 29, 2022, 8:52am UTC](https://discuss.pynq.io/t/how-can-i-use-my-custom-dataset-on-pynq-z2-for-qnn-which-is-already-trained-on-pascal-voc-dataset/4612/2 "2022-08-29T08:52:19Z")

</div>

Welcome,

Please find the MNIST dataset loading method as referencing.

> [@MNIST + AXI Stream on PYNQ 2.7 (Attention to Details)](https://discuss.pynq.io/t/mnist-axi-stream-on-pynq-2-7-attention-to-details/4267):
>
> Hello PYNQ community: I wish you all doing great =] Previous support posts are having trouble on AXI stream HLS about CNN design and DMA: So I am going to share the details on 2020.2 HLS tool and PYNQ 2.7 design. The Github Repository Link for this tutorial: HLS AXI Stream People keep asking about tlast is missing, which this is common that after version (not sure when) the structure C no longer interpret as we expected: Old struct AXI\_DMA\_IO{ ap\_int\<16\> data; ap\_int\<1\>last; }; New #in…
