PyTorch Image Object Detection Faster R-CNN ResNet-50 Batch GPU

Model: PyTorch Image Object Detection — Faster R-CNN ResNet-50 FPN (pretrained on COCO) Accelerator: Tesla T4 GPU (fixed batch size) Host: 50 × n1-standard-4 (4 vCPUs, 15 GB RAM)

This batch pipeline performs object detection using an open-source PyTorch Faster R-CNN ResNet-50 FPN model on GPU. It reads image URIs from GCS, decodes and preprocesses images, and runs batched inference with a fixed batch size to measure stable GPU performance.

The following graphs show various metrics when running PyTorch Image Object Detection Faster R-CNN ResNet-50 Batch GPU pipeline. See the glossary for definitions.

Full pipeline implementation is available here.

What is the estimated cost to run the pipeline?

RunTime and EstimatedCost

How has various metrics changed when running the pipeline for different Beam SDK versions?

AvgThroughputBytesPerSec by Version

AvgThroughputElementsPerSec by Version

How has various metrics changed over time when running the pipeline?

AvgThroughputBytesPerSec by Date

AvgThroughputElementsPerSec by Date