|
|
--- |
|
|
language: |
|
|
- en |
|
|
metrics: |
|
|
- accuracy |
|
|
pipeline_tag: image-classification |
|
|
tags: |
|
|
- arxiv:1512.03385 |
|
|
--- |
|
|
|
|
|
# Resnet50-v1.5 |
|
|
|
|
|
- Origin: https://github.com/mlcommons/inference/tree/master/vision/classification_and_detection |
|
|
- Note : https://catalog.ngc.nvidia.com/orgs/nvidia/resources/resnet_50_v1_5_for_pytorch |
|
|
|
|
|
## Model |
|
|
|
|
|
- ONNX attributes: |
|
|
- opset : 11 |
|
|
- Operators: 233 (11 types) |
|
|
- Add: 17 |
|
|
- ArgMax: 1 |
|
|
- Conv2D: 36 |
|
|
- MatMul: 1 |
|
|
- PaddedConv2D: 17 |
|
|
- PaddedMaxPooling2D: 1 |
|
|
- Producer: 108 |
|
|
- ReLU: 49 |
|
|
- ReduceMean: 1 |
|
|
- Softmax: 1 |
|
|
- Squeeze: 1 |
|
|
|
|
|
## Aidge support |
|
|
|
|
|
> Note: We tested this network for the following features. If you encounter any error please open an [issue](https://gitlab.eclipse.org/groups/eclipse/aidge/-/issues). Features not tested in CI may not be functional. |
|
|
|
|
|
| Feature | Tested in CI | |
|
|
| :---------: | :----------: | |
|
|
| ONNX import | ✔ | |
|
|
| Backend CPU | ✔ | |
|
|
|
|
|
## ImageNet1k |
|
|
- Input: |
|
|
- size: \[N, 3, 224, 224\] |
|
|
- format : NCHW |
|
|
- Output |
|
|
- size: [N, 1000] |
