Update README.md
Browse files
README.md
CHANGED
|
@@ -23,13 +23,13 @@ The model is trained on 2000 examples for each color and shape combo (64,000 sam
|
|
| 23 |
The model is tested/evaluated on the dataset [https://huggingface.co/datasets/sdtemple/colored-shapes](https://huggingface.co/datasets/sdtemple/colored-shapes) which has slightly smaller shapes simulated (out of distribution) relative to the training data.
|
| 24 |
|
| 25 |
- Accuracy: 75%
|
| 26 |
-
- Min precision (triangle):
|
| 27 |
-
- Max precision (rectangle):
|
| 28 |
-
- Min recall (diamond):
|
| 29 |
-
- Max recall (triangle):
|
| 30 |
-
- AUROC (macro-averaged):
|
| 31 |
- Min AUROC (diamond): 90%
|
| 32 |
-
- Max AUROC (circle):
|
| 33 |
|
| 34 |
Compared to [https://huggingface.co/sdtemple/color-prediction-model](https://huggingface.co/sdtemple/color-prediction-model), it is harder to predict the shape than the color of the object.
|
| 35 |
|
|
|
|
| 23 |
The model is tested/evaluated on the dataset [https://huggingface.co/datasets/sdtemple/colored-shapes](https://huggingface.co/datasets/sdtemple/colored-shapes) which has slightly smaller shapes simulated (out of distribution) relative to the training data.
|
| 24 |
|
| 25 |
- Accuracy: 75%
|
| 26 |
+
- Min precision (triangle): 57%
|
| 27 |
+
- Max precision (rectangle): 98%
|
| 28 |
+
- Min recall (diamond): 66%
|
| 29 |
+
- Max recall (triangle): 84%
|
| 30 |
+
- AUROC (macro-averaged): 92%
|
| 31 |
- Min AUROC (diamond): 90%
|
| 32 |
+
- Max AUROC (circle): 94%
|
| 33 |
|
| 34 |
Compared to [https://huggingface.co/sdtemple/color-prediction-model](https://huggingface.co/sdtemple/color-prediction-model), it is harder to predict the shape than the color of the object.
|
| 35 |
|