Gradient° has been updated in response to a ton of feedback from the community. Here's a roundup of some of the things we've added recently:
New System & Custom Metrics!
Now whenever you run a job you can watch in real-time the host metrics including GPU Utilization (%), GPU Memory Used, Temperature, and Load. You can also see CPU and RAM utilization as well to ensure that your Gradient job is performing as expected.
Additionally, you can create your own Custom Metrics. Read more in the docs.
Build Dockerfiles on the fly
Give us a Dockerfile and pass the
--useDockerfile command to the CLI and we will build the image before running it! Can also optionally push to a public or private docker registry.
Learn more in the docs. Or clone a sample job into your own account!
New sample project - StyleGAN face generator!
You might have seen the buzz around https://thispersondoesnotexist.com. Behind the scenes it is built on NVIDIA's revolutionary new StyleGAN.
You can test out Gradient jobs with the Job Builder and the StyleGAN sample project in just one click.
Here's a public job showing the output of the GAN: https://www.paperspace.com/console/jobs/jqtl5zat16r6/metrics
SAML Single Sign-On (SSO)
It is now easier than ever to integrate Gradient with your existing enterprise deployment. Learn more here.
New RAPIDS Notebook Container
www.rapids.ai is an exciting new project to bring the power of the GPU to the traditional (non-deep-learning) machine learning and data science world. With NVIDIA Rapids, you can execute data science and analytics pipelines on GPUs.