Once you've uploaded a model, it becomes available on the platform within the model registry. You can start sending one-off requests through the model, or running your datasets through to get insights. Start by clicking on the model to open the model overview page, which houses both the inference playground and the linked datasets.

Inference Playground

The inference playground is a great way to start getting your hands in the mud. Here, you can send your own requests through the model to see what and how it thinks.
The inference playground takes a minute or so to spin up after first deploying the model. Also, the first request may take longer than subsequent requests due to cold start.

Linked Datasets

When you've uploaded both a model and it's corresponding dataset (let's say a test or validation set), you still need to conceptually "link" the two. When you create a task with chosen model(s) and dataset(s), the models and datasets are linked automatically. We'll get to this more in the "Tasks" section of this guide:
Linking a dataset and a model allows you to run the dataset through the model and get both high-level metrics (standard accuracy, F1, precision, and recall) as well as row-level insights. More on this in the "Run Reports" section:
Last modified 4mo ago