> For the complete documentation index, see [llms.txt](https://docs.seaart.ai/guide-1/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.seaart.ai/guide-1/seaart-ai-basic-function/lora-training.md).
# LoRA Training
Train, edit, and publish a custom LoRA model on SeaArt. Let's dive in!
## Page entry
You can open LoRA training in two ways:
* Click **Models** in the left navigation, then click **Online Training**.
* Click **Create** in the top navigation bar, select **Model**, then choose **Train**.
## How to Train a LoRA
\*For detailed steps and advanced training parameter settings, click here.
{% content-ref url="/pages/bDLwl9DD9y2ga18oYqz4" %}
[LoRA Training (Advance)](/guide-1/advanced-guide/lora-training-advance.md)
{% endcontent-ref %}
## Publish a LoRA
When training is complete, click **Publish** at the bottom of the completed training page. This opens the publish editor, where you set the model details, version, and images before publishing.
The editor has three steps: **Edit Model**, **Edit Version**, and **Add Image**. Use **Next** to continue and **Save** on the last step.
### Step 1: Edit Model
Enter the model information:
* **Model/LORA Name** - required.
* **Model Type** - set to LoRA.
* **Set Tags** - optional tags that help others find the model.
* **Model Introduction** - required description of the model.
* **Source** - choose Original or Reprint.
* **Public or Private** - Private is selected by default.
Set user permissions:
* **Allow Online Use on SeaArt**
* **Allow Downloads**
* **Allow Model Merging**
Set the commercial license:
* **Generated content can be sold or used for commercial purposes**
* **Resale of Models or Their Sale After Merging**
Original models can also be published through the [Creator Incentive Program](https://docs.seaart.ai/guide-1/permanent-events/seaart.ai-creator-incentive-program).
Click **Next**.
### Step 2: Edit Version
Fill in the version details:
* **Version Name** - required, such as v1.0.
* **Base Model** - required; select the base model used for training.
* **Model Used** - select the model associated with this version.
* **Support for "Train Similar Models"** - optional toggle.
* **Version Introduction** - describe this version.
* **Select a File** - upload the .safetensors file.
* **Trigger Words** - optional. Enable "This version does not require trigger words" if the model does not use them.
Review or adjust the training and recommended parameters shown for this version, including Epochs, Iteration Steps, Strength, and Clip Skip.
Use **Add Version** if you need more than one version, or **Toggle Version Info** to show or hide version details.
Click **Next**.
### Step 3: Add Image
Add showcase images and a cover for the model:
* **Version Works** - required; upload 1–20 images or select relevant works.
* **Cover image** - click **Change Cover** to choose the image shown on the model page.
Do not upload images containing explicit, violent, gory, or other inappropriate content.
Click **Save** to publish the LoRA with the settings you entered.
## Edit a Published LoRA
To edit a model that has already been published, open its model detail page. Click the three-dot menu in the top-right corner and select **Edit**. This opens the same three-step editor.
## View Trained LoRAs
Open [Model Training](https://www.seaart.ai/model-training) to view the LoRAs you have trained. From this page, you can edit a trained LoRA or start another training run.
