Skip to content

Plan an AI Toolkit training workload

Check training hardware and persistence before using an AI Toolkit image.

Published on Oct 12, 2026 | Product: GPU

Before you start

AI Toolkit training can require 24 GB or more of VRAM. Only use templates exposed as compatible for the selected GPU and provider; do not treat an available Docker image as a certified purchase option.

Steps

  1. Check the model, training method, dataset permissions, and minimum GPU memory.
  2. Choose a compatible offering with enough VRAM and disk for base weights, datasets, checkpoints, and temporary files.
  3. Keep datasets and outputs on the persistent workspace and save a reproducible configuration.
  4. Start with a short validation run using the image’s documented training command.
  5. Monitor memory and checkpoints. Export checkpoints before stopping or destroying the rental.

Check the result

A validation run completes on supported hardware before you commit to a long training job.

Troubleshooting

A 12 GB GPU cannot certify workflows that require 24 GB. If no compatible template is offered, contact support about availability rather than bypassing the compatibility check.

Last updated October 12, 2026. Availability and prices are shown in the console before checkout.
Still need help? Open a support ticket with your service name, the time of the issue, and a redacted error message. Never include passwords or API keys.
Categories
Billing and support CLI and developer tools DocVault AI GPU Cloud GPU applications Getting started Hosting and DNS Managed AI APIs Object Storage VPS
Products
DocVault AI GPU Object Storage Shared Hosting VPS