Cerebras Wafer-Scale Cluster Documentation (R1.8)#
Setup Cerebras virtual environment with the latest Cerebras software release.
Clone Cerebras Model Zoo with examples of transformers and computer vision models.
Launch your job using Cerebras scripts for PyTorch and TensorFlow.
Cerebras job scheduling and monitoring looking at the queue of the Cerebras cluster.
Explore the components of Cerebras Wafer-Scale Cluster and their role during training of neural networks.
Understand Weight Streaming and Pipelined Execution, the two modes that neural networks can be trained and evaluated using the Cerebras cluster.
Find commonly used terms in the Glossary.
Navigate Ways to port your model and leverage the reosurces of Cerebras Model Zoo.
Prepare your data using offline Python scripts and online dataloaders.
Choose one of thre different apporaches to port your model: Port your PyTorch code using Cerebras Model Zoo, Port your code using Cerebras PyTorch API (Experimental), or Port your TensorFlow code using Cerebras Model Zoo.
Find the list of Cerebras Model Zoo Supported Operations API.
Find the list of Cerebras Model Zoo YAML params.
In these sections, you will find additional tools available in the Cerebras Wafer-Scale Cluster and Cerebras Model Zoo, as well as common errors and workarounds.