.. _cs-tf-workflow:

Workflow for TensorFlow on CS
=============================

When you are using TensorFlow targeting Cerebras CS system, start with the high-level workflow described here.

.. raw:: html

    <div class="container-fluid">

    <div class="row">
        <div class="col">
            <div class="card h-100" >
                <div class="card-body">
                    <h3 class="card-title">1</h3>
                    <a href="#port-to-cerebras" class="card-link"><h4>Port to Cerebras</h4></a>
                </div>
            </div>
        </div>
        <div class="col">
            <div class="card h-100" >
                <div class="card-body">
                    <h3 class="card-title">2</h3>

                    <a href="#prepare-input" class="card-link"><h4>Prepare input</h4></a>
                </div>
            </div>
        </div>
        <div class="col">
            <div class="card h-100" >
                <div class="card-body">
                    <h3 class="card-title">3</h3>

                    <a href="#compile-on-cpu" class="card-link"><h4>Compile on CPU</h4></a>
                </div>
            </div>
        </div>

        <!-- <div class="col">
            <div class="card h-100" >
                <div class="card-body">
                    <h3 class="card-title">4</h3>
                    <p class="card-text">Estimate Slurm resources</p>
                    <a href="running-a-model/cs-input-analyzer.html" class="card-link">Use cs_input_analyzer script</a>
                </div>
            </div>
        </div> -->

        <div class="col">
            <div class="card h-100" >
                <div class="card-body">
                    <h3 class="card-title">4</h3>

                    <a href="#run-the-job-on-cs-system" class="card-link"><h4>Run on Cerebras</h4></a>
                </div>
            </div>
        </div>

        <!-- break -->

    </div>
    </div>
    <br />

.. _port-short:

Port to Cerebras
----------------

Prepare your model and the input function by using ``CerebrasEstimator`` in place of the TensorFlow Estimator.

To run on the CS system you must use ``CerebrasEstimator``. However, when you use ``CerebrasEstimator`` you can also run on a CPU or GPU with minimal code change.

See :ref:`porting-tf-to-cs` to accomplish this step.

.. _prep-input-short:

Prepare input
-------------

To achieve a high input data throughput often requires that your input pipeline is running on multiple worker processes across many CPU nodes at once. Use this :ref:`preparing-tf-input` documentation to organize your input data pipeline.

.. attention::

	Wherever you place your model and the associated scripts in your Original Cerebras Support-Cluster, ensure that these are accessible from every node in the cluster.

.. _compile-on-cpu-short:

Compile on CPU
--------------

Before you run it on the CS system, we recommend that you iterate until your model first compiles successfully on a CPU node that has the Cerebras Singularity container client software. To accomplish this step, proceed as follows:

- First run in the ``validate_only``	mode.
- Then run full compile with ``compile_only``.

See this :ref:`validate-and-compile-on-cpu` to perform this step.

	.. note::

			The :ref:`run-py-template` documents an example template that can help you in organizing your code so these steps can become easy to manage. However, the ``run.py`` described in :ref:`run-py-template` is an example template only. You can organize your code whichever way that best suits you, as long as you use ``CerebrasEstimator``.

    .. admonition:: Hardware requirements

				Also make sure that the hardware you use for this step satisfies the minimum requirements stated in :ref:`hw-requirements-compile-only`.

.. _run-on-cs-short:

Run the job on CS system
------------------------

Finally, train, eval or predict on the CS system. Use the checkpoints to do evaluation on a GPU or a CPU. See this :ref:`train-eval-predict` for documentation on this step.

..
    4. Estimate the performance and the needed resources
        Use the ``cs_input_analyzer`` to estimate the performance and Slurm resource settings for your ``input_fn`` and model. See :ref:`cs-input-analyzer` for how to use the ``cs_input_analyzer`` tool.