A collection of functions to help you easily train and run Tensorflow Keras
Project description
Keras Toolkit
A collection of functions to help you easily train and run Tensorflow Keras
Quickstart
Install the library:
pip install keras-toolkit
You can now use it:
import keras_toolkit as kt
# kt reduces the number of lines from ~100 to ~3
strategy = kt.accelerator.auto_select(verbose=True)
decoder = kt.image.build_decoder(with_labels=True, target_size=(300, 300))
dtrain = kt.data.build_dataset(paths, labels, bsize=BATCH_SIZE, decode_fn=decoder)
with strategy.scope():
model = tf.keras.Sequential([...])
model.compile(...)
model.fit(...)
API References
Get the complete API reference here.
Acknowledgement
The auto_select_accelerator
was inspired by Martin Gorner's Kaggle notebook.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
keras-toolkit-0.1.0rc1.tar.gz
(4.1 kB
view hashes)
Built Distribution
Close
Hashes for keras_toolkit-0.1.0rc1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | df0e010ef2de21f93cd5a542bdd8f9bef210c9b9f3e8020597bfe8ba14604438 |
|
MD5 | 55ac0829b58590def541c97bb9e0eeb1 |
|
BLAKE2b-256 | 02d39d25caa841dfd57ca36e0cb62a3f3dd6487cc249dbce6b7135a401cf904a |