File size: 1,329 Bytes
27a7410 fb91e68 b9bed0f 0e823a9 27a7410 e4c50f5 a328010 27a7410 7e28732 27a7410 a328010 27a7410 e4c50f5 daff781 a328010 daff781 27a7410 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 |
---
library_name: tf-keras
tags:
- computer-vision
- image-classification
---
## Model description
This repo contains the trained model Self-supervised contrastive learning with SimSiam on CIFAR-10 Dataset.
Keras link: https://keras.io/examples/vision/simsiam/
Full credits to https://twitter.com/RisingSayak
## Intended uses & limitations
The trained model can be used as a learned representation for downstream tasks like image classification.
## Training and evaluation data
The dataset we are using here is called CIFAR-100. The CIFAR-10 dataset consists of 60000 32x32 color images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images.
Two particular augmentation transforms that seem to matter the most are:
- Random resized crops
- Color distortions
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
| name | learning_rate | decay | momentum | nesterov | training_precision |
|----|-------------|-----|--------|--------|------------------|
|SGD|{'class_name': 'CosineDecay', 'config': {'initial_learning_rate': 0.03, 'decay_steps': 3900, 'alpha': 0.0, 'name': None}}|0.0|0.8999999761581421|False|float32|
## Model Plot
<details>
<summary>View Model Plot</summary>
</details> |