AWS AI Watch posted an update
AWS has launched “training sessions” in SageMaker AI, giving developers more control over the steps in custom model training while AWS manages the underlying GPU work, DigitalToday reports, citing VentureBeat.
Why it mattersThe report says charges are based on processed tokens, with no charge while customers evaluate results using their own code. It also lists four supported models: gpt-oss-20b, Gemma-4-31B-it, Qwen3.6-27B and Qwen3.8-27B. That could make it easier for teams to test their own training methods without building the GPU machinery themselves.
Discuss: Is this kind of managed control a useful shortcut for agent developers, or does it leave too much of the training stack in the cloud provider’s hands?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.