Define Models

AI models are a core resource for inferencing. The project manifest allows you to easily define models whether hosted by Hypermode or another host.

hypermode.json

{
  "models": [
    {
      "name": "sentiment-classifier",
      "task": "classification",
      "sourceModel": "distilbert/distilbert-base-uncased-finetuned-sst-2-english",
      "provider": "hugging-face",
      "host": "hypermode"
    }
  ]
}

Properties

name · string · required

Internal name of your AI model. Used for indicating the model to use in an API call.

task · string · required

Training intent of model, such as classification or embedding.

sourceModel · string · required

Relative path of model within provider.

provider · string · required

Source location of model, such as hugging-face or openai.

host · string · required

Runtime provider of model.

Show host types

hypermode · string

Dynamically deployed models on Hypermode.

custom · string

Fine-tuned models on Hypermode.

... · string

Name of a project-defined host.

Auto-deployed models

When using hugging-face as the provider and hypermode as the host, Hypermode automatically deploys a dedicated instance of the defined sourceModel when deploying your project. Your project’s functions securely connect to the hosted model, with no further configuration required.