Train on the cloud¶
Audience: Users who want to develop and train models on the cloud (public cloud, private cloud or onprem clusters).
1: Access a GPU machine on the cloud
Learn to train models using an interactive cloud machine.
2: Run a model in the background on the cloud
Learn to run a model in the background on a cloud machine.
3: Save up to 80% on cloud costs with fault-tolerant training
Run on the cloud for 1/10th the price with fault-tolerant training.
4: Run many models at once
Run many models at once (sweep) to find the best performing model.
Grid.ai is the official cloud training solution for PyTorch Lightning. Grid is designed to support researcher workloads at both academic labs and major companies.