|
|
--- |
|
|
datasets: |
|
|
- uoft-cs/cifar10 |
|
|
pipeline_tag: image-classification |
|
|
--- |
|
|
|
|
|
https://github.com/mlcommons/tiny/blob/master/benchmark/training/image_classification/trained_models/pretrainedResnet.tflite |
|
|
|
|
|
# Resnet8 |
|
|
|
|
|
- Origin: https://github.com/mlcommons/tiny/blob/master/benchmark/training/image_classification/ |
|
|
|
|
|
## Model |
|
|
|
|
|
- ONNX attributes: |
|
|
- opset : 13 |
|
|
- Operators: 45 (10 types) |
|
|
- Add: 3 |
|
|
- Conv2D: 2 |
|
|
- PaddedConv2D: 7 |
|
|
- Producer: 21 |
|
|
- ReLU: 7 |
|
|
- AveragePooling: 1 |
|
|
- Transpose: 1 |
|
|
- Reshape: 1 |
|
|
- FC: 1 |
|
|
- Softmax: 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 | ✔️ | |
|
|
| Runtime CPU | ✔️ | |
|
|
| Runtime CUDA | ✔️ | |
|
|
| Export CPU | ✔️ | |
|
|
|
|
|
## CIFAR10 |
|
|
- Input: |
|
|
- size: \[N, 3, 32, 32\] |
|
|
- format : NCHW |
|
|
- Output |
|
|
- size: [N, 10] |
