MLable

TensorFlow utilities for layers, metrics, shaping, sampling, and model building blocks.

Installation

pip install -U mlable

Quickstart

import tensorflow as tf
import mlable.layers.shaping
import mlable.metrics

x = tf.ones(shape=(2, 4, 8))
y = mlable.layers.shaping.Divide(axis=-1, factor=4, insert=True)(x)

metric = mlable.metrics.CategoricalGroupAccuracy(group=4)
score = metric(y_true=tf.one_hot([[1, 2, 3, 4]], depth=8), y_pred=tf.one_hot([[1, 2, 0, 4]], depth=8))

Documentation

Detailed usage documentation is available under docs/sources/:

• docs/sources/layers/shaping.md
• docs/sources/layers/embedding.md
• docs/sources/layers/transformer.md
• docs/sources/metrics/group_accuracy.md
• docs/sources/index.md

Credits

Andrej Karpathy reconnected my ML synapses with micrograd.

License

Licensed under the aGPLv3.