ixrt

Wide ResNet50 (ixRT)

Model Description

The distinguishing feature of Wide ResNet50 lies in its widened architecture compared to traditional ResNet models. By increasing the width of the residual blocks, Wide ResNet50 enhances the capacity of the network to capture richer and more diverse feature representations, leading to improved performance on various visual recognition tasks.

Supported Environments

GPU	IXUCA SDK	Release	Branch
MR-V100	4.4.0	26.03	release/26.03
MR-V100	4.3.0	25.12	release/25.12

Note: 请切换到与您的 SDK 版本对应的 Release 分支进行测试。请勿直接在 master 分支上运行测试，因为 master 分支可能包含与您的本地 SDK 版本不兼容的最新更改。

切换分支命令示例：git checkout release/26.03

Model Preparation

Prepare Resources

Pretrained model: https://download.pytorch.org/models/wide_resnet50_2-95faca4d.pth

Dataset: https://www.image-net.org/download.php to download the validation dataset.

Install Dependencies

# Install libGL
## CentOS
yum install -y mesa-libGL
## Ubuntu
apt install -y libgl1-mesa-glx

pip3 install -r ../../ixrt_common/requirements.txt

Model Conversion

mkdir -p checkpoints/
python3 ../../ixrt_common/export.py --model-name wide_resnet50_2 --weight wide_resnet50_2-95faca4d.pth --output checkpoints/wide_resnet50.onnx

Model Inference

export PROJ_DIR=./
export DATASETS_DIR=/path/to/imagenet_val/
export CHECKPOINTS_DIR=./checkpoints
export RUN_DIR=../../ixrt_common/
export CONFIG_DIR=../../ixrt_common/config/WIDE_RESNET50_CONFIG

FP16

# Accuracy
bash scripts/infer_wide_resnet50_fp16_accuracy.sh
# Performance
bash scripts/infer_wide_resnet50_fp16_performance.sh

INT8

# Accuracy
bash scripts/infer_wide_resnet50_int8_accuracy.sh
# Performance
bash scripts/infer_wide_resnet50_int8_performance.sh

Model Results

Model	BatchSize	Precision	FPS	Top-1(%)	Top-5(%)
Wide ResNet50	32	FP16	2478.551	78.486	94.084
Wide ResNet50	32	INT8	5981.702	76.956	93.920