Ray Train Benchmarks#
Below we document key performance benchmarks for common Ray Train tasks and workflows.
GPU image training#
This task uses the TorchTrainer module to train different amounts of data using an Pytorch ResNet model.
We test out the performance across different cluster sizes and data sizes.
Note
For multi-host distributed training, on AWS we need to ensure ec2 instances are in the same VPC and all ports are open in the secure group.
Cluster Setup |
Data Size |
Performance |
Command |
1 g3.8xlarge node (1 worker) |
1 GB (1623 images) |
79.76 s (2 epochs, 40.7 images/sec) |
|
1 g3.8xlarge node (1 worker) |
20 GB (32460 images) |
1388.33 s (2 epochs, 46.76 images/sec) |
|
4 g3.16xlarge nodes (16 workers) |
100 GB (162300 images) |
434.95 s (2 epochs, 746.29 images/sec) |
|
Pytorch Training Parity#
This task checks the performance parity between native Pytorch Distributed and Ray Train’s distributed TorchTrainer.
We demonstrate that the performance is similar (within 2.5%) between the two frameworks. Performance may vary greatly across different model, hardware, and cluster configurations.
The reported times are for the raw training times. There is an unreported constant setup overhead of a few seconds for both methods that is negligible for longer training runs.
Cluster Setup |
Dataset |
Performance |
Command |
4 m5.2xlarge nodes (4 workers) |
FashionMNIST |
196.64 s (vs 194.90 s Pytorch) |
|
4 m5.2xlarge nodes (16 workers) |
FashionMNIST |
430.88 s (vs 475.97 s Pytorch) |
|
4 g4dn.12xlarge node (16 workers) |
FashionMNIST |
149.80 s (vs 146.46 s Pytorch) |
|
Tensorflow Training Parity#
This task checks the performance parity between native Tensorflow Distributed and Ray Train’s distributed TensorflowTrainer.
We demonstrate that the performance is similar (within 1%) between the two frameworks. Performance may vary greatly across different model, hardware, and cluster configurations.
The reported times are for the raw training times. There is an unreported constant setup overhead of a few seconds for both methods that is negligible for longer training runs.
Note
The batch size and number of epochs is different for the GPU benchmark, resulting in a longer runtime.
Cluster Setup |
Dataset |
Performance |
Command |
4 m5.2xlarge nodes (4 workers) |
FashionMNIST |
78.81 s (vs 79.67 s Tensorflow) |
|
4 m5.2xlarge nodes (16 workers) |
FashionMNIST |
64.57 s (vs 67.45 s Tensorflow) |
|
4 g4dn.12xlarge node (16 workers) |
FashionMNIST |
465.16 s (vs 461.74 s Tensorflow) |
|
XGBoost training#
This task uses the XGBoostTrainer module to train on different sizes of data with different amounts of parallelism to show near-linear scaling from distributed data parallelism.
XGBoost parameters were kept as defaults for xgboost==1.7.6
this task.
Cluster Setup |
Number of distributed training workers |
Data Size |
Performance |
Command |
1 m5.4xlarge node with 16 CPUs |
1 training worker using 12 CPUs, leaving 4 CPUs for Ray Data tasks |
10 GB (26M rows) |
310.22 s |
|
10 m5.4xlarge nodes |
10 training workers (one per node), using 10x12 CPUs, leaving 10x4 CPUs for Ray Data tasks |
100 GB (260M rows) |
326.86 s |
|