---
license: other
tags:
- generated_from_keras_callback
model-index:
- name: Xanadu00/galaxy_classifier
  results: []
---

<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->

# Xanadu00/galaxy_classifier_mobilevit

This model is a fine-tuned version of [apple/mobilevit-small](https://huggingface.co/apple/mobilevit-small) on [Galaxy10 DECals dataset](https://astronn.readthedocs.io/en/latest/galaxy10.html#).
It achieves the following results on the evaluation set:
- Train Loss: 0.3550
- Train Accuracy: 0.8797
- Validation Loss: 0.5428
- Validation Accuracy: 0.8326
- Epoch: 14

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

### Dataset Fields

The dataset has the following fields (also called "features"):

```json
{
  "image": "Image(decode=True, id=None)",
  "target": "ClassLabel(names=['Barred Spiral', 'Cigar Round Smooth', 'Distributed', 'Edge-on with Bulge', 'Edge-on without Bulge', 'In-between Round Smooth', 'Merging', 'Round Smooth', 'Unbarred Loss Spiral', 'Unbarred Tight Spiral'], id=['0', '1', '2', '3', '4', '5', '6', '7', '8', '9'])"
}
```

### Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

| Split name   | Num samples         |
| ------------ | ------------------- |
| train        | 14188 |
| valid        | 3548 |

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- optimizer: {'name': 'AdamW', 'weight_decay': 0.004, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 0.002, 'decay_steps': 10000, 'end_learning_rate': 2e-05, 'power': 0.5, 'cycle': False, 'name': None}}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
- training_precision: float32

### Training results

| Train Loss | Train Accuracy | Validation Loss | Validation Accuracy | Epoch |
|:----------:|:--------------:|:---------------:|:-------------------:|:-----:|
| 0.9179     | 0.6833         | 0.7849          | 0.7342              | 0     |
| 0.6763     | 0.7702         | 0.7576          | 0.7556              | 1     |
| 0.6042     | 0.7912         | 0.5977          | 0.7909              | 2     |
| 0.5604     | 0.8063         | 0.5499          | 0.8117              | 3     |
| 0.5176     | 0.8213         | 0.5738          | 0.8114              | 4     |
| 0.4944     | 0.8353         | 0.5317          | 0.8176              | 5     |
| 0.4735     | 0.8368         | 0.6051          | 0.7914              | 6     |
| 0.4427     | 0.8483         | 0.5357          | 0.8275              | 7     |
| 0.4430     | 0.8480         | 0.5327          | 0.8250              | 8     |
| 0.4127     | 0.8596         | 0.4947          | 0.8362              | 9     |
| 0.4017     | 0.8603         | 0.5378          | 0.8162              | 10    |
| 0.3840     | 0.8694         | 0.5070          | 0.8326              | 11    |
| 0.3677     | 0.8722         | 0.4875          | 0.8295              | 12    |
| 0.3606     | 0.8778         | 0.5071          | 0.8360              | 13    |
| 0.3550     | 0.8797         | 0.5428          | 0.8326              | 14    |


### Framework versions

- Transformers 4.30.2
- TensorFlow 2.12.0
- Datasets 2.13.1
- Tokenizers 0.13.3