Every week, hundreds of new shiny new ML algorithms go live and by the time you clone it, set up environments and start replicating it, there is a new and better algorithm.
That's why we built "Recipes", a ready-to-run curated ML Algorithm library.
How recipes help your ML team
Your ML team no longer needs to waste time finding the best algorithm for your use case. With recipes, you can start working without worrying about setting up the code, environment, and compute.
Pick any algorithm and start working on it in a few clicks.
What else we have been working on
Granular usage billing
You can now download a detailed cost statement and get resource utilisation reports for each instance and datastore.
Heavy lifting GPUs
We have beefed up the compute power available to our users with a 4 x Nvidia T4 machine and more CPUs.
Cluster cost estimator
Forecast how much every cluster might cost so that you are no longer surprised by your AWS bill.
We added accelerator tags to nodes so now you can tag instances with their accelerator types - giving Kubernetes context to select the right node and avoid errors.