Created
November 14, 2018 21:27
-
-
Save jlebar/f3bb841f1d3e03ca7abe17b851fefec3 to your computer and use it in GitHub Desktop.
Run resnet50 benchmark on GCE
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gcloud compute ssh $INSTANCE_NAME | |
# Clone TensorFlow benchmark repository. | |
git clone https://github.com/tensorflow/benchmarks.git && cd benchmarks | |
git reset --hard 1e7d788042dfc6d5e5cd87410c57d5eccee5c664 | |
cd scripts/tf_cnn_benchmarks | |
## Synthetic data test | |
# 8 GPUs | |
python tf_cnn_benchmarks.py \ | |
--batch_size=364 \ | |
--num_batches=100 \ | |
--model=resnet50 \ | |
--optimizer=momentum \ | |
--variable_update=replicated \ | |
--all_reduce_spec=nccl \ | |
--use_fp16=True \ | |
--nodistortions \ | |
--gradient_repacking=2 \ | |
--compute_lr_on_cpu=True \ | |
--single_l2_loss_op=True \ | |
--xla_compile=True \ | |
--num_gpus=8 \ | |
--loss_type_to_report=base_loss | |
# 1 GPU | |
python tf_cnn_benchmarks.py \ | |
--batch_size=364 \ | |
--num_batches=100 \ | |
--model=resnet50 \ | |
--optimizer=momentum \ | |
--use_fp16=True \ | |
--nodistortions \ | |
--compute_lr_on_cpu=True \ | |
--single_l2_loss_op=True \ | |
--xla_compile=True \ | |
--loss_type_to_report=base_loss | |
## Real data test | |
# add --data_dir=/data/imagenet to the 1 or 8 GPU command. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment