.. raw:: html

<div align="center">
  <img src="docs/source/_static/logo.png"><br><br>
</div>

.. image:: https://img.shields.io/badge/python-3.6+-blue.svg :alt: Supported Python versions

.. image:: https://img.shields.io/badge/tensorflow-2.7.2+-orange.svg :alt: Supported TensorFlow versions

.. image:: https://codecov.io/gh/eliorc/tavolo/branch/master/graph/badge.svg :target: https://codecov.io/gh/eliorc/tavolo :alt: Code test coverage

Tavolo

| tavolo_ aims to package together valuable modules and functionality written for TensorFlow_ high-level Keras API for ease of use. | You see, the deep learning world is moving fast, and new ideas keep on coming. | tavolo gathers implementations of these useful ideas from the community (by contribution, from Kaggle_ etc.) and makes them accessible in a single PyPI hosted package that compliments the tf.keras_ module.

Documentation

| Tavolo documentation_

.. _Tavolo documentation: https://tavolo.readthedocs.io/

Showcase

| tavolo's API is straightforward and adopting its modules is as easy as it gets. | In tavolo, you'll find implementations for basic like PositionalEncoding_ and non-layer implementations that can ease development, like the LearningRateFinder_. | For example, if we wanted to add head a Yang-style attention mechanism into our model and look for the optimal learning rate, it would look something like:

.. code-block:: python3

import tensorflow as tf
import tavolo as tvl

model = tf.keras.Sequential([
    tf.keras.layers.Embedding(input_dim=vocab_size, output_dim=embedding_size, input_length=max_len),
    tvl.seq2vec.YangAttention(n_units=64),  # <--- Add Yang style attention
    tf.keras.layers.Dense(n_hidden_units, activation='relu'),
    tf.keras.layers.Dense(1, activation='sigmoid')])

model.compile(optimizer=tf.keras.optimizers.SGD(), loss=tf.keras.losses.BinaryCrossentropy())

# Run learning rate range test
lr_finder = tvl.learning.LearningRateFinder(model=model)

learning_rates, losses = lr_finder.scan(train_data, train_labels, min_lr=0.0001, max_lr=1.0, batch_size=128)

### Plot the results to choose your learning rate

.. _tavolo: https://github.com/eliorc/tavolo .. _TensorFlow: https://www.tensorflow.org/ .. _Kaggle: https://www.kaggle.com .. _tf.keras: https://www.tensorflow.org/guide/keras .. _PositionalEncoding: https://tavolo.readthedocs.io/en/latest/embeddings.html#module-embeddings.PositionalEncoding .. _MultiHeadedAttention: https://tavolo.readthedocs.io/en/latest/seq2seq.html#multi-headed-self-attention .. _LearningRateFinder: https://tavolo.readthedocs.io/en/latest/learning.html#learning-rate-finder

Contributing

| Want to contribute? Please read our Contributing guide_.

.. _Contributing guide: https://tavolo.readthedocs.io/en/latest/contributing.html