> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.seelab.ai/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Managing a model: details, retraining, deletion

Your Seelab models evolve with your products. From the **Custom models** library, you can rename a model, view its images, retrain it with new photos or delete it.

![The Custom models library](https://storage.crisp.chat/users/helpdesk/website/-/6/8/0/3/6803ea00be5f5800/image_1647jl7.jpg =1000xauto)

# The library

Each model card shows its type, its creation date and its author, with two buttons:

* **Use model**: opens Creation with this model selected.
* **Edit**: opens the model page.

# The model page

The model page has three tabs.

* **Parameters**: the model's name, type and thumbnail.
* **History**: the model's history.
* **Dataset**: the model's training images.

![A model page, Dataset tab](https://storage.crisp.chat/users/helpdesk/website/-/6/8/0/3/6803ea00be5f5800/image_1k44r4n.jpg =1000xauto)

# Retrain a model

Use **Retrain** when:

* your product has changed (new label, new packaging);
* a detail is poorly rendered and you have better photos to show it;
* you want to enrich the model with new angles.

From the **Dataset** tab, add or remove images, then click **Retrain**. Each retraining counts against your **Remaining Monthlies**, shown in the **Trainings** panel on the model page.

# Share a model

A model belongs to the project it was created in: every project member can see it and use it. To make it available to another team, invite its members to the project.

# Delete a model

Deleting a model requires confirmation.

||| Deletion is permanent: the model can't be recovered, and it disappears for every project member.

# Glossary

* **Model page**: the page that brings together a model's parameters, history and dataset.
* **Retrain**: run the training of an existing model again with new images.
* **Project**: a team's shared space; its members have access to the same models.