# Test and validate model provider settings

After configuring the model settings in the `Provider settings` tab, you can verify your configuration by clicking `Test / Cache model`. This will send a completion request to the selected model provider and display the results in a dialog, as shown below:

<figure><img src="https://1757931399-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FeqMpFKstQ48aOht1u50p%2Fuploads%2Fi9n7HTUJNNFdrBemMP1B%2Fimage.png?alt=media&#x26;token=0c5b6e67-5995-4f73-8860-d0e7d0183521" alt="" width="366"><figcaption></figcaption></figure>

When using the `Local / Custom` model provider, BurpGPT will automatically download and cache the specified model—if it is a valid publicly accessible [Hugging Face model](https://huggingface.co/models)—before displaying the result dialog. An example of this process is shown in the screenshot below. Subsequent queries will use the cached model, avoiding repeated downloads.

<figure><img src="https://1757931399-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FeqMpFKstQ48aOht1u50p%2Fuploads%2Fctrc3fiZBegMxHPEkOKt%2Fllm_server_caching_light.png?alt=media&#x26;token=cb9ba4d3-7cc3-4772-b592-68b74db4e329" alt=""><figcaption><p>Model download and caching.</p></figcaption></figure>

{% hint style="info" %}
Please keep in mind that the caching process can take a considerable amount of time, depending on the model size and the speed of your internet connection. Additionally, ensure that your file system has sufficient free space to accommodate the downloaded models.
{% endhint %}

Downloaded models are saved in the cache folder at these locations:

* Linux and Mac: `~/.cache/huggingface/transformers/`
* Windows: `%userprofile%/.cache/huggingface/transformers/`

{% hint style="info" %}
These cached models can later serve as a foundation for training your own custom models—a topic we may cover in a future post.
{% endhint %}