Cerebras AI Model Studio Documentation#
The Cerebras AI Model Studio is a computing service powered by dedicated clusters of Cerebras CS-2s and hosted by Cirrascale Cloud Services. It is a purpose-built platform, optimized for training large language models on dedicated clusters of millions of cores. It provides deterministic performance, requires no distributed computing headaches, and is push-button simple to start.
Interested? Sign up to get started!
To train your models in the Cerebras AI Model Studio, you can use one of these methods:
Model Studio Launchpad
Model Studio Launchpad is a pay-by-the-model service powered by a simple CLI. You can train from scratch:
- GPT-3 1.3B, 6.7B, and 13B
- GPT-J and GPT NeoX
- Eleuther GPT-J, Eleuther GPT-NeoX
- CodeGen 350M, 2.7B, 6.1B, and 16.1B parameters
- Cerebras-GPT 111M, 256M, 590M, 1.3B, 2.7B, 6.7B, and 13B parameters
Model Studio Builder
Model Studio Builder gives you full access to a Cerebras Wafer-Scale Cluster and Cerebras Model Zoo to further customize your model.
More information about the Cerebras Model Zoo interface in the Cerebras Wafer-Scale Cluster documentation.