Create Dataset
Creating a dataset is the first step in using cloud services. You can think of it as needing storage space for cleaning data, and the dataset is that storage space for the data.
Last updated
Creating a dataset is the first step in using cloud services. You can think of it as needing storage space for cleaning data, and the dataset is that storage space for the data.
Last updated
On the Supametas.AI main page, you can view your existing datasets. Click on " + Click to create a dataset " to enter the setup wizard:
Before starting the setup, please ensure that you have your multimodal large model API ready. Currently, we support OpenAI's multimodal model, OpenAI-gpt4o. To obtain the API, visit https://platform.openai.com/api-keys or use a third-party OpenAI service provider.
Enter Dataset Name:
Create a name for the new dataset within the input box, limited to 20 characters.
Enter Dataset Dntroduction(optional):
Add a brief description for the new dataset, limited to 50 characters.
Select Model Type:
In the "Model Configuration" section, choose between using the system's built-in model or configuring and using an external model.
Once the quota for the built-in model is used up, it will no longer be selectable.
We recommend configuring an external model.
Input API Configuration: If you opt for "Configure and use external models", follow these steps for configuration:
API KEY: Enter your API key.
BaseUrl: Input the API's BaseUrl.
Channel: Select the API channel from the dropdown list, for example, OpenAI.
Model: Choose the specific model version from the dropdown list, like gpt-4-turbo.
Save and Continue: After completing all settings, click the "Next" button at the bottom of the page to save settings and proceed.
Upon saving, the system will automatically validate the API configuration. If incorrect, you'll receive prompts to re-enter; if correct, a dataset will be created automatically.
Note: The dataset currently uses the configured API model. You can use different API models each time you create or modify a dataset to control usage and costs.
Built-in model quotas are limited, so we recommend using your own API or contacting us via email to customize quotas at service@supametas.ai.