Model hubs · HTTP
Hugging Face Hub
How do I use Hugging Face Hub with Onrup?
Publish a finished run to a public or private Hub repository in one call. A model card is generated from the run — base model, dataset reference, hyperparameters, final metrics, GPU class and wall time — and you can edit it before or after.
The model hub the open-weight ecosystem is organised around, for both base models and published fine-tunes.
Minimal working example
curl https://api.onrup.com/v1/runs/$RUN_ID/publications \
-H "Authorization: Bearer $ONRUP_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"repo_id": "your-org/support-triage-v3",
"private": true,
"include_metrics": true
}'The thing that catches people out
Publishing is never automatic and is not charged for. Nothing leaves your tenant until you make this call, and datasets are never included — only weights and the card.
Worth knowing
- Every base model in the catalogue is identified by its Hub repository, so what you fine-tune is exactly what the Hub publishes.
- A private repository is a portability mechanism: the model exists somewhere we do not control.
- The base model’s licence governs the derivative, and the generated card records which licence applies.
Start with the free tier
A magic link creates your account, your tenant and your first API key. No card until you ask for compute.