Scaling to trillion-parameter model training on AWS - Amazon Science

Retrieved on: 2022-09-26 19:13:09

Tags for this article:

Machine learning

Artificial intelligence

Computing

Artificial neural networks

Deep learning

Computational neuroscience

Computational statistics

Cloud infrastructure

PyTorch

Graphics processing unit

Backpropagation

Amazon Web Services

Click the tags to see associated articles and topics

Scaling to trillion-parameter model training on AWS - Amazon Science. View article details on HISWAI: https://www.amazon.science/blog/scaling-to-trillion-parameter-model-training-on-aws

Excerpt

AWS recently announced a preview of Amazon EC2 P4de GPU instances powered by 400 Gbps networking and 80GB GPU memory. P4de provides twice as much ...

Article found on: www.amazon.science

View Original Article