tensorflow/lite/g3doc/api_docs/python/tflite_model_maker/text_classifier/MobileBertClassifierSpec.md
page_type: reference description: Creates MobileBert model spec for the text classification task. See also: <a href="../../tflite_model_maker/text_classifier/BertClassifierSpec"><code>tflite_model_maker.text_classifier.BertClassifierSpec</code></a>.
<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="tflite_model_maker.text_classifier.MobileBertClassifierSpec" /> <meta itemprop="path" content="Stable" /> </div>Creates MobileBert model spec for the text classification task. See also: <a href="../../tflite_model_maker/text_classifier/BertClassifierSpec"><code>tflite_model_maker.text_classifier.BertClassifierSpec</code></a>.
<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>tflite_model_maker.text_classifier.MobileBertClassifierSpec( *, uri='https://tfhub.dev/google/mobilebert/uncased_L-24_H-128_B-512_A-4_F-4_OPT/1', model_dir=None, seq_len=128, dropout_rate=0.1, initializer_range=0.02, learning_rate=3e-05, distribution_strategy='off', num_gpus=-1, tpu='', trainable=True, do_lower_case=True, is_tf2=False, name='MobileBert', tflite_input_name=None, default_batch_size=48, index_to_label=None ) </code></pre> <!-- Placeholder for "Used in" --> <!-- 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> `uri`<a id="uri"></a> </td> <td> TF-Hub path/url to Bert module. </td> </tr><tr> <td> `model_dir`<a id="model_dir"></a> </td> <td> The location of the model checkpoint files. </td> </tr><tr> <td> `seq_len`<a id="seq_len"></a> </td> <td> Length of the sequence to feed into the model. </td> </tr><tr> <td> `dropout_rate`<a id="dropout_rate"></a> </td> <td> The rate for dropout. </td> </tr><tr> <td> `initializer_range`<a id="initializer_range"></a> </td> <td> The stdev of the truncated_normal_initializer for initializing all weight matrices. </td> </tr><tr> <td> `learning_rate`<a id="learning_rate"></a> </td> <td> The initial learning rate for Adam. </td> </tr><tr> <td> `distribution_strategy`<a id="distribution_strategy"></a> </td> <td> A string specifying which distribution strategy to use. Accepted values are 'off', 'one_device', 'mirrored', 'parameter_server', 'multi_worker_mirrored', and 'tpu' -- case insensitive. 'off' means not to use Distribution Strategy; 'tpu' means to use TPUStrategy using `tpu_address`. </td> </tr><tr> <td> `num_gpus`<a id="num_gpus"></a> </td> <td> How many GPUs to use at each worker with the DistributionStrategies API. The default is -1, which means utilize all available GPUs. </td> </tr><tr> <td> `tpu`<a id="tpu"></a> </td> <td> TPU address to connect to. </td> </tr><tr> <td> `trainable`<a id="trainable"></a> </td> <td> boolean, whether pretrain layer is trainable. </td> </tr><tr> <td> `do_lower_case`<a id="do_lower_case"></a> </td> <td> boolean, whether to lower case the input text. Should be True for uncased models and False for cased models. </td> </tr><tr> <td> `is_tf2`<a id="is_tf2"></a> </td> <td> boolean, whether the hub module is in TensorFlow 2.x format. </td> </tr><tr> <td> `name`<a id="name"></a> </td> <td> The name of the object. </td> </tr><tr> <td> `tflite_input_name`<a id="tflite_input_name"></a> </td> <td> Dict, input names for the TFLite model. </td> </tr><tr> <td> `default_batch_size`<a id="default_batch_size"></a> </td> <td> Default batch size for training. </td> </tr><tr> <td> `index_to_label`<a id="index_to_label"></a> </td> <td> List of labels in the training data. e.g. ['neg', 'pos']. </td> </tr> </table>