Rigging the Lottery: Making All Tickets Winners

Paper: https://arxiv.org/abs/1911.11134

Best Sparse Models

Parameters are float, so each parameter is represented with 4 bytes. Uniform sparsity distribution keeps first layer dense therefore have slightly larger size and parameters. ERK applies to all layers except for 99% sparse model, in which we set the first layer to be dense, since otherwise we observe much worse performance.

Extended Training Results

Performance of RigL increases significantly with extended training iterations. In this section we extend the training of sparse models by 5x. Note that sparse models require much less FLOPs per training iteration and therefore most of the extended trainings cost less FLOPs than baseline dense training.

With using only 2.32x of the original FLOPS we can train 99% sparse Resnet-50 that obtains an impressive 66.94% test accuracy.

S. Distribution	Sparsity	Training FLOPs	Inference FLOPs	Model Size (Bytes)	Top-1 Acc	Ckpt
- (DENSE)	0	3.2e18	8.2e9	102.122	76.8	-
ERK	0.8	2.09x	0.42x	23.683	77.17	link
Uniform	0.8	1.14x	0.23x	23.685	76.71	link
ERK	0.9	1.23x	0.24x	13.499	76.42	link
Uniform	0.9	0.66x	0.13x	13.532	75.73	link
ERK	0.95	0.63x	0.12x	8.399	74.63	link
Uniform	0.95	0.42x	0.08x	8.433	73.22	link
ERK	0.965	0.45x	0.09x	6.904	72.77	link
Uniform	0.965	0.34x	0.07x	6.904	71.31	link
ERK	0.99	0.29x	0.05x	4.354	61.86	link
ERK	0.99	2.32x	0.05x	4.354	66.94	link

1x Training Results

S. Distribution	Sparsity	Training FLOPs	Inference FLOPs	Model Size (Bytes)	Top-1 Acc	Ckpt
ERK	0.8	0.42x	0.42x	23.683	75.12	link
Uniform	0.8	0.23x	0.23x	23.685	74.60	link
ERK	0.9	0.24x	0.24x	13.499	73.07	link
Uniform	0.9	0.13x	0.13x	13.532	72.02	link

Sparse Training Algorithms

In this repository we implement following dynamic sparsity strategies:

SET: Implements Sparse Evalutionary Training (SET) which corresponds to replacing low magnitude connections randomly with new ones.
SNFS: Implements momentum based training without sparsity re-distribution:
RigL: Our method, RigL, removes a fraction of connections based on weight magnitudes and activates new ones using instantaneous gradient information.

And the following one-shot pruning algorithm:

SNIP: Single-shot Network Pruning based on connection sensitivity prunes the least salient connections before training.

We have code for following settings:

Imagenet2012: TPU compatible code with Resnet-50 and MobileNet-v1/v2.
CIFAR-10 with WideResNets.
MNIST with 2 layer fully connected network.

Setup

First clone this repo.

git clone https://github.com/google-research/rigl.git
cd rigl

We use Neurips 2019 MicroNet Challenge code for counting operations and size of our networks. Let's clone the google_research repo and add current folder to the python path.

git clone https://github.com/google-research/google-research.git
mv google-research/ google_research/
export PYTHONPATH=$PYTHONPATH:$PWD

Now we can run some tests. Following script creates a virtual environment and installs the necessary libraries. Finally, it runs few tests.

bash run.sh

We need to activate the virtual environment before running an experiment. With that, we are ready to run some trivial MNIST experiments.

source env/bin/activate

python rigl/mnist/mnist_train_eval.py

You can load and verify the performance of the Resnet-50 checkpoints like following.

python rigl/imagenet_resnet/imagenet_train_eval.py --mode=eval_once --training_method=baseline --eval_batch_size=100 --output_dir=/path/to/folder --eval_once_ckpt_prefix=s80_model.ckpt-1280000 --use_folder_stub=False

We use the Official TPU Code for loading ImageNet data. First clone the tensorflow/tpu repo and then add models/ folder to the python path.

git clone https://github.com/tensorflow/tpu.git
export PYTHONPATH=$PYTHONPATH:$PWD/tpu/models/

Disclaimer

This is not an official Google product.

sunxh16/rigl