This repository holds NVIDIA-maintained utilities to improve GPU performance and usability for Tensorflow training and inference. The intent of Atex is to make up-to-date utilities available to users as quickly as possible. As such, these utilities are experimental. Some we will upstream and support permanently in projects such as TensorFlow or Keras. Others will eventually be discontinued.
nv_norms.LayerNormalization and nv_norms.InstanceNormalization are fused
implementations designed to replace the tf.keras.layers.LayerNormalization and
tfa.layers.InstanceNormalization. The Keras and Addons implementations compose
many ops, such as Mul, Add, etc., to compute the stats and then scale/offset the
results. In contrast, the NVIDIA fused implementation provided in nv_norms
compute the norms in a single operation, eliminating many expensive round-trips
to device memory and significantly improving performance. More in
nv_ops.
Please review the Contribution Guidelines.
GitHub issues will be used for tracking requests and bugs.