page_type: reference description: Representative dataset used to optimize the model.

<link rel="stylesheet" href="/site-assets/css/style.css"> <!-- DO NOT EDIT! Automatically generated file. --> <div itemscope itemtype="http://developers.google.com/ReferenceObject"> <meta itemprop="name" content="tf.lite.RepresentativeDataset" /> <meta itemprop="path" content="Stable" /> <meta itemprop="property" content="__init__"/> </div>

tf.lite.RepresentativeDataset

<!-- Insert buttons and diff --> <table class="tfo-notebook-buttons tfo-api nocontent" align="left"> <td> <a target="_blank" href="https://github.com/tensorflow/tensorflow/blob/v2.11.0/tensorflow/lite/python/lite.py#L158-L179"> </a> </td> </table>

Representative dataset used to optimize the model.

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>tf.lite.RepresentativeDataset( input_gen ) </code></pre> <!-- Placeholder for "Used in" -->

This is a generator function that provides a small dataset to calibrate or estimate the range, i.e, (min, max) of all floating-point arrays in the model (such as model input, activation outputs of intermediate layers, and model output) for quantization. Usually, this is a small subset of a few hundred samples randomly chosen, in no particular order, from the training or evaluation dataset.

<!-- Tabular view --> <table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2"><h2 class="add-link">Args</h2></th></tr> <tr> <td> `input_gen`<a id="input_gen"></a> </td> <td> A generator function that generates input samples for the model and has the same order, type and shape as the inputs to the model. Usually, this is a small subset of a few hundred samples randomly chosen, in no particular order, from the training or evaluation dataset. </td> </tr> </table>