Aidan N. Gomez, Ivan Zhang, Kevin Swersky, Yarin Gal, and Geoffrey E. Hinton

• Requirements
• Quick Start
• Experiments

• Python 3
• Tensorflow 1.8

1. Train a model: python -m TD.train --hparams=resnet_default
2. Prune that model: python -m TD.scripts.prune.eval --hparams=resnet_default --prune_percent 0.0,0.25,0.5,0.75,0.95

• --env: one of local, gcp (GPU instances), or tpu (TPU instances). Feel free to add more if necessary.
• --hparams: the hparam set you want to run.
• --hparam_override: manually specify hparams to be overridden (e.g --hparam_override 'drop_rate=0.66')