Skip to main content
This feature is in Private Preview. For access or more information, contact us or reach out to your account representative.
Upload fine-tuned models from your S3 bucket, track multiple versions, and deploy them to your dedicated endpoint. For detailed endpoint documentation, see the Management API Reference.

​
Key Concepts

  • Model Architecture: The underlying structure of a model (e.g., gpt-oss-120b, llama3.1-8b). Your dedicated endpoint is provisioned for a specific architecture, and any custom weights you upload must be compatible with that architecture.
  • Model Version: Each time you upload custom weights, a new version is created. Versions are identified by an auto-incrementing integer (e.g., 1, 2, 3) and can also have user-defined aliases like production or v1-stable.
  • Endpoint: Your dedicated endpoint is identified by a unique ID (e.g., my-org-gpt-oss-120b). This ID is used as the model field when making inference requests.

​
Typical Workflow

  1. Upload model weights — Push your fine-tuned weights from S3 to Cerebras. The upload is asynchronous; you’ll receive a version ID to track progress.
  2. Check upload status — Poll the version status until the sync completes.
  3. Deploy to endpoint — Once the upload is complete, deploy the version to your dedicated endpoint.
  4. Run inference — Make requests to your endpoint using the same endpoint ID as the model field.
  5. Iterate — Upload new versions as you fine-tune, assign aliases like production to track releases, and deploy updates with zero downtime.

​
Authentication

The Management API uses a separate API key from the standard inference API. You can find your Management API key under Management API keys on the API keys page in the Cerebras Cloud console.

​
S3 Bucket Setup

Before uploading model weights, you need an S3 bucket configured with cross-account access to Cerebras. Your Cerebras representative will provide specific instructions, but the bucket policy must grant cross-account access: 
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "AWS": "<cerebras-provided-iam-role-arn>"
      },
      "Action": [
        "s3:GetObject",
        "s3:GetObjectVersion",
        "s3:ListBucket",
        "s3:GetBucketLocation"
      ],
      "Resource": [
        "arn:aws:s3:::<your-bucket-name>",
        "arn:aws:s3:::<your-bucket-name>/*"
      ]
    }
  ]
}
Replace <your-bucket-name> with your S3 bucket name. The IAM role ARN (<cerebras-provided-iam-role-arn>) will be provided by Cerebras and enables secure cross-account access to your model weights.