Enjoy the "big ferocious" performance of NVIDIA's Ampere-based RTX 3090 for less than $1 an hour. Supplies are limited so reserve one while you can.

The trainML platform has been extended to support batch inference jobs, enabling customers to use trainML for all stages of the machine learning pipeline that require GPU acceleration.

The trainML platform now allows customers to store models permanently and reuse those models for as many notebook and training jobs as desired.

You can now view summary details of the contents of a created dataset from the user interface.

In addition to centrally viewing job and dataset log output in real-time, you can now download the full log extract for datasets and job workers after they have finished.

New job environments based on Python 3.8 are now available for all frameworks.

​trainML jobs and datasets can now be created, managed, and monitored programmatically using our Python CLI/SDK.

You can now convert notebooks directly into training jobs to easily run independent training experiments while working on your projects. In contrast to copying the notebook into another notebook job, training jobs will run autonomously, send their output to the location you specify, and automatically terminate when finished.

​trainML notebooks can now be forked into new instances to enable easy parallel experimentation. Unlike other cloud notebooks, when you fork a trainML notebook, the entire working directory is copied. All datasets, checkpoints, and other data are copied into the new notebook.