Skip to main content

Model Repository

TIR Model Repositories are designed to store model weights and configuration files. These repositories can be backed by either E2E Object Storage (EOS) or PVC storage within a Kubernetes environment.

Model Repository Resource​

Example Usage​

 resource "tir_model_repository" "jn"{
    name = "soss"
    storage_type = "new"
    model_type = "pytorch"
    team_id = <team_id : string>
    project_id = <project_id:string>
    active_iam = <active_iam:string>
  }

Schema​

Required​

  • active_iam (String) : The IAM (Identity and Access Management) role associated with the model repository.
  • model_type (String) : The type of model stored in the repository. This defines the category or framework of the model (e.g., TensorFlow, PyTorch).
  • name (String) : The name of the model repository. This is a required field and must be unique.
  • project_id (String) : The ID of the project where the model repository is deployed.
  • storage_type (String) : The type of storage for the model repository. Supported values are 'new' for managed storage, 'existing' for E2E S3, and 'external' for PVC (Persistent Volume Claim). If you are choosing storage_type other than "new" please check optional fields there are some other required fields too. Otherwise resources will not created.
  • team_id (String) : The ID of the team that owns the model repository.

Optional​

  • access_key (String) : The access key for the model repository. This is required incase of storage_type as external.
  • bucket_name (String) : The name of the bucket associated with the model repository. This is required incase of storage_type as existing or external
  • secret_key (String) : The secret key for the model repository. This is required incase of storage_type as external

Read-Only​

  • bucket_endpoint (String) : The endpoint URL for accessing the bucket. This is computed automatically.
  • bucket_url (String) : The URL of the bucket associated with the model repository. This is computed automatically.
  • created_at (String) : The timestamp when the model repository was created. This is computed automatically.
  • id (String) : The ID of this resource.
  • status (String) : The current status of the model repository. This is computed automatically.

Supported Model Type​

  • pytorch
  • triton
  • ensorrt
  • custom