ResNet-50: Optimized for AMD ROCm

ResNet-50 is a 50-layer deep residual network for image classification over the 1000 ImageNet categories. This repository packages inference for image classification using ONNX Runtime, exported and validated for AMD ROCm so it runs efficiently on AMD GPUs, CPUs, and NPUs.

This is based on the implementation of ResNet-50 found here. This repository contains configurations and scripts optimized for AMD® ROCm™ platforms. You can use the resnet-50 AMD scripts to reproduce results or export with custom configurations.

---

Task Overview

Task: Image classification

Dataset: ImageNet-1000 label space (a small sample set is staged under dataset/samples/ for visual evaluation)

Output metrics: Throughput (inferences/sec), latency (mean/P95/P99 ms), per-operator profiling breakdown

---

AMD ROCm Optimization

This model export has been adapted and validated for AMD Instinct™ / Radeon™ GPUs running ROCm, as well as AMD CPUs and AMD Ryzen AI NPUs. Key points:

• Validated backends: ONNX Runtime across CPU (native execution provider), GPU (MIGraphX execution provider, ROCm-based), and NPU (VitisAI execution provider).
• No code changes required versus the upstream microsoft/resnet-50 export — only environment/runtime configuration differs.
• CPU fallback path supported for environments without a ROCm-capable GPU.

Runtime	Precision	Backend	Hardware	Notes
ONNX Runtime	FP32 / FP16 / BF16 / INT8	CPU Execution Provider	AMD CPU	—
ONNX Runtime	FP32 / FP16 / BF16 / INT8	MIGraphX Execution Provider	AMD Instinct™ / Radeon™ GPU (ROCm)	—
ONNX Runtime	FP32	VitisAI Execution Provider	AMD Ryzen AI NPU	Accepts FP32 input; VitisAI quantizes internally

---

Getting Started

For setup instructions, evaluation scripts, and custom configuration options, see the resnet-50 on GitHub.

---

Model Details

Model Type: Image classification, ResNet-50 (50-layer deep residual network)

Base Model: microsoft/resnet-50 (ResNet-50, ImageNet-1000)

Model Stats:

• Input resolution: 224×224 (pixel_values: (batch, 3, 224, 224); logits output: (batch, 1000))
• Precision tested: FP32, FP16, BF16, INT8 (CPU/GPU); FP32 auto-quantized (NPU)

---

Accuracy Pipeline

Accuracy evaluation is not yet implemented for this model.

---

Dig Deeper

Want to explore the full evaluation scripts, config options, and other AMD-optimized model examples?

📂 View the full project on GitHub

The GitHub repository includes:

• Benchmark and profile scripts for CPU, GPU, and NPU (including AI Analyzer profiling)
• ONNX export and pretrained weight download automation
• Sample-image evaluation with annotated top-prediction overlays
• make metrics aggregation into METRICS_TABLE.md