ResNeXt50 is a simple, highly modularized network architecture for image classification, constructed by a template with cardinality = 32 and bottleneck width = 4d and defined by Xie et al. in their paper.
| Model | Download | Top-1 accuracy (%) | Top-5 accuracy (%) |
|---|---|---|---|
| ResNeXt50 | 96 MB | 81.096 | 95.326 |
- ResNeXt50 model is from the paper titled Aggregated Residual Transformations for Deep Neural Networks.
- This onnx model is converted from a pytorch model, which is pretrained on ImageNet by ©RossWightman with the code in his repository.
Apache 2.0