| | --- |
| | library_name: tf-keras |
| | --- |
| | |
| | ## Model description |
| | **This model is implementation of the distillation recipe proposed in DeiT.** |
| | Visit Keras example on [Distilling Vision Transformers](https://keras.io/examples/vision/deit/). |
| | |
| | Full credits to: [Sayak Paul](https://twitter.com/RisingSayak) |
| | |
| | In the original Vision Transformers (ViT) paper (Dosovitskiy et al.), the authors concluded that to perform on par with Convolutional Neural Networks (CNNs), ViTs need to be pre-trained on larger datasets. The larger the better. This is mainly due to the lack of inductive biases in the ViT architecture -- unlike CNNs, they don't have layers that exploit locality. |
| | |
| | Many groups have proposed different ways to deal with the problem of data-intensiveness of ViT training. One such way was shown in the Data-efficient image Transformers, (DeiT) paper (Touvron et al.). The authors introduced a distillation technique that is specific to transformer-based vision models. DeiT is among the first works to show that it's possible to train ViTs well without using larger datasets. |
| |
|
| | ## Intended uses & limitations |
| |
|
| | The model is trained for demonstrative purposes and does not guarantee the best results in production. |
| | For better results, follow & optimize the [Keras example](https://keras.io/examples/vision/deit/) as per your need. |
| |
|
| | ## Training and evaluation data |
| |
|
| | The model is trained and evaluated on [TF Flowers dataset](https://www.tensorflow.org/datasets/catalog/tf_flowers) |
| |
|
| | ## Training procedure |
| |
|
| | Training procedure is followed exactly as from the [keras example](https://keras.io/examples/vision/deit/). |
| | The batch size is however decreased to 16 from the original 256 for accomodating the model in a single V100 GPU memory. |
| |
|
| | ### Training hyperparameters |
| |
|
| | The following hyperparameters were used during training: |
| |
|
| | | name | learning_rate | decay | beta_1 | beta_2 | epsilon | amsgrad | weight_decay | exclude_from_weight_decay | training_precision | |
| | |----|-------------|-----|------|------|-------|-------|------------|-------------------------|------------------| |
| | |AdamW|6.25000029685907e-05|0.0|0.8999999761581421|0.9990000128746033|1e-07|False|9.999999747378752e-05|None|float32| |
| |
|
| | ## Model Plot |
| |
|
| | <details> |
| | <summary>View Model Plot</summary> |
| |
|
| |  |
| |
|
| | </details> |
