Spell has partnered with Omdena to support their AI for Good mission in helping organizations build AI models for social good and solve real-world social problems through AI.
Spell is a cloud platform for building and managing data science and machine learning projects. Spell takes care of the setup of infrastructure required for machine learning projects, which makes machine learning projects easier to start and faster to get results. Spell was developed to make experiments faster to get started and Omdena participants can take advantage of free GPU credits and the Spell support team.
The fastest and most powerful end-to-end platform for data science and machine learning.
Free up your local CPU machine and do your Data Analysis in the cloud. Spin up CPU or GPU powered Jupyter Workspaces in seconds and work on data prep, cleaning, transformation and experimentation seamlessly and efficiently. Keep information organized in a flexible development environment for Jupyter notebooks, code, and data.
Powerful CPU and GPU compute
Spell offers easy access to the fastest hardware available to train models in the cloud. This includes NVidia K80, P100 to V100x8 and more. Utilize Spot Instances in one-click and reduce experimentation costs by up to 90%. Furthermore, the platform makes it simple to distribute code to run across a number of processes and across multiple machines simultaneously.
Open-source ML frameworks at your fingertips
Spell environments are pre-built with the most popular open-source frameworks used for Machine Learning and Deep Learning. Tensorflow, PyTorch, Caffe, Fast.ai, MXNet and many others are supported. Save time on setup and managing versions and focus on experimentation.
Share your Omdena and Spell experience and knowledge of data science and machine learning with others and grow in the AI community. By writing a blog/article that talks about how you used Spell in your process and linking back to us, you could be eligible for $250 of GPU credit. Get in contact to learn more firstname.lastname@example.org