Skip to content

Commit

Permalink
Add tensorflow resnet152 imagenet
Browse files Browse the repository at this point in the history
  • Loading branch information
Wang committed Dec 1, 2017
1 parent 4786014 commit 9693021
Showing 1 changed file with 21 additions and 0 deletions.
21 changes: 21 additions & 0 deletions task_config_template.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -60,3 +60,24 @@ metrics = ['Images per sec', 'Top 1 accuracy']
compute_method = ['average', 'last']
command_to_execute = python tensorflow_benchmark/tf_cnn_benchmarks/tf_cnn_benchmarks.py --data_name=cifar10 --num_gpus=8 --batch_size=32 --model=resnet50 --variable_update=parameter_server --print_training_accuracy=True
num_gpus = 8

[tensorflow_resnet152_p3_2xlg]
patterns = ['images/sec: (\d+\.\d+)', 'top_1_accuracy: (\d+\.\d+)']
metrics = ['Images per sec', 'Top 1 accuracy']
compute_method = ['average', 'last']
command_to_execute = python tensorflow_benchmark/tf_cnn_benchmarks/tf_cnn_benchmarks.py --data_name=imagenet --num_gpus=1 --batch_size=32 --model=resnet152 --variable_update=parameter_server --print_training_accuracy=True
num_gpus = 1

[tensorflow_resnet152_p3_8xlg]
patterns = ['images/sec: (\d+\.\d+)', 'top_1_accuracy: (\d+\.\d+)']
metrics = ['Images per sec', 'Top 1 accuracy']
compute_method = ['average', 'last']
command_to_execute = python tensorflow_benchmark/tf_cnn_benchmarks/tf_cnn_benchmarks.py --data_name=imagenet --num_gpus=4 --batch_size=32 --model=resnet152 --variable_update=parameter_server --print_training_accuracy=True
num_gpus = 4

[tensorflow_resnet152_p3_16xlg]
patterns = ['images/sec: (\d+\.\d+)', 'top_1_accuracy: (\d+\.\d+)']
metrics = ['Images per sec', 'Top 1 accuracy']
compute_method = ['average', 'last']
command_to_execute = python tensorflow_benchmark/tf_cnn_benchmarks/tf_cnn_benchmarks.py --data_name=imagenet --num_gpus=8 --batch_size=32 --model=resnet152 --variable_update=parameter_server --print_training_accuracy=True
num_gpus = 8

0 comments on commit 9693021

Please sign in to comment.