cerebras.modelzoo.common#

`checkpoint_utils`	Various utility functions related to checkpoint saving and loading.
`dump_context`	Provides DumpContext, a debug utility for dumping activations and gradients on a CPU/GPU run, and setting up debug names for dumped WSE activations to be automatically correlated.
`eval_harness_impl`	Implements CSEvalHarnessAdapter class for running Eval Harness on CSX via the cstorch API
`half_dtype`	Module which provides utilities for selecting 16-bit floating point representation.
`input`
`input_utils`
`mlops`
`model_utils`
`pytorch_utils`	General purpose Pytorch Utilities
`registry`	This is registry for the cerebras modelzoo
`registry_cli`
`run_cstorch_eval_harness`	Eval Harness run script
`run_cstorch_flow`	Generic run scripts build using the cstorch API
`run_gpt_inference`	GPT Inference script built using the cstorch API
`run_utils`	Utilities for running Cerebras Pytorch Models
`train_and_eval`	Generic run scripts build using the cstorch API
`utils`

cerebras.modelzoo

cerebras.modelzoo.common.checkpoint_utils