Update README.md
Browse files
README.md
CHANGED
|
@@ -39,18 +39,30 @@ It achieves the following results on the evaluation set:
|
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
| 42 |
-
|
| 43 |
|
| 44 |
## Intended uses & limitations
|
| 45 |
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
## Training and evaluation data
|
| 49 |
|
| 50 |
-
|
|
|
|
|
|
|
|
|
|
| 51 |
|
| 52 |
## Training procedure
|
| 53 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
### Training hyperparameters
|
| 55 |
|
| 56 |
The following hyperparameters were used during training:
|
|
|
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
| 42 |
+
This model is a fine-tuned version of Google's vit-base-patch16-224-in21k designed to identify geological hand samples.
|
| 43 |
|
| 44 |
## Intended uses & limitations
|
| 45 |
|
| 46 |
+
Currently the VIT is fine-tuned on 10 classes:
|
| 47 |
+
|
| 48 |
+
['Andesite', 'Basalt', 'Chalk', 'Dolomite', 'Flint', 'Gneiss', 'Granite', 'Limestone', 'Sandstone', 'Slate']
|
| 49 |
+
|
| 50 |
+
Future iteartions of the model will feature an expanded breadth of rock categories.
|
| 51 |
+
|
| 52 |
|
| 53 |
## Training and evaluation data
|
| 54 |
|
| 55 |
+
The model performs relatively well on 10 classes of rocks - with some confusion between limestone and other carbonates.
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
|
| 59 |
|
| 60 |
## Training procedure
|
| 61 |
|
| 62 |
+
495 images of geological hand samples were selected with an 80:20 train-test/validation split.
|
| 63 |
+
|
| 64 |
+
Classes were roughly equally represented across the 495 samples.
|
| 65 |
+
|
| 66 |
### Training hyperparameters
|
| 67 |
|
| 68 |
The following hyperparameters were used during training:
|