Original Cerebras Installation Documentation (R1.6.1)

Run your first Cerebras job

• Setup your environment

• Clone Cerebras Model Zoo with examples of models implementations

• Launch your job using Cerebras scripts for PyTorch and TensorFlow

How Cerebras works

• Explore the components of the Original Cerebras Installation

• Understand Pipelined Execution

• Find commonly used terms in the Glossary

Port your model

• Navigate Ways to port your model and leverage the reosurces of Cerebras Model Zoo

• Prepare your data using offline Python scripts and online dataloaders

• Wrap your PyTorch code to run on Cerebras

• Wrap your TensorFlow code to run on Cerebras

Cerebras features

• The csrun_cpu Script

• The csrun_wse Script

• The cs_input_analyzer Script

• Mixed-Precision Training

• Dynamic Loss Scaling

• Performance Optimization Practices

• Multi-Replica Data Parallel Training

• Training with Sparse input

Compiler reports

• Input Function Report

• Compile Report

• Incremental Compile

Extensions

• Adding Custom Packages To cbcore Container

previous

Glossary

next

Software Release Notes