Overview
Turn your own documents and examples into a fine-tuned model your application can call, on GPUs in your own RunPod account.
Welcome to Tensorant. You bring your documents or examples, review them in the console, train a model on GPUs in your own RunPod account, and call it from your application like any OpenAI model.
Where to start
- New here? What Tensorant is gives you the whole picture in a few minutes.
- Want to try it? The Quickstart takes you from sign-up to your first API call.
- Setting up your organization? Connections shows what to copy from RunPod and Hugging Face.
- Calling a model that is already published? Go straight to the API quickstart.
What's in the docs
- Get started: What Tensorant is, Quickstart, Accounts, organizations and roles, Connections.
- Data: Projects, Sources, Vision models, Recipes, Review, Versions.
- Train: Experiments and Training runs.
- Evaluate: Evaluations.
- Serve: Playground, Deployments, Publishing models, Serving settings, Publishing to Hugging Face.
- Use the API: API quickstart, Chat completions, Keys and usage.
- Organization: Members and invitations, Plans and billing, Activity and the audit trail.
- Reference: Supported models, Limits, Troubleshooting.
Getting help
When the console refuses something, its message usually says what to do next, and Troubleshooting covers the common ones. For anything else, write to [email protected].