Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: en
|
| 3 |
+
license: mit
|
| 4 |
+
datasets: [Cats-Dogs]
|
| 5 |
+
metrics: [accuracy, f1, precision, recall]
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# Transfer Learning On Alexnet:
|
| 9 |
+
|
| 10 |
+
This model uses Transfer Learning for a 2 labelled class Cats-Dogs classification.
|
| 11 |
+
|
| 12 |
+
- Pre-trained model used: Alexnet.
|
| 13 |
+
- Dataset used: Cats_Dogs.
|
| 14 |
+
- batch_size = 8
|
| 15 |
+
- learning_rate = 0.001
|
| 16 |
+
- epochs = 5
|
| 17 |
+
|
| 18 |
+
## **Evaluation Results**
|
| 19 |
+
|
| 20 |
+
### 📌 **Before Transfer Learning **
|
| 21 |
+
**Accuracy:** 40.66%
|
| 22 |
+
- Precision: 0.3983,
|
| 23 |
+
- Recall: 0.4066,
|
| 24 |
+
- F1-score: 0.3942
|
| 25 |
+
Confusion Matrix:
|
| 26 |
+
[[ 659 1841]
|
| 27 |
+
[1126 1374]]
|
| 28 |
+
---
|
| 29 |
+
|
| 30 |
+
### ✅ **After Training**
|
| 31 |
+
|
| 32 |
+
** Train Accuracy:** 92.40%
|
| 33 |
+
- Precision: 0.9250,
|
| 34 |
+
- Recall: 0.9240,
|
| 35 |
+
- F1-score: 0.9240
|
| 36 |
+
Confusion Matrix:
|
| 37 |
+
[[9486 514]
|
| 38 |
+
[1006 8994]]
|
| 39 |
+
Training Finished.
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
|
| 43 |
+
** Validatio Accuracy:** 94.10%
|
| 44 |
+
- Precision: 0.9425,
|
| 45 |
+
- Recall: 0.9410,
|
| 46 |
+
- F1-score: 0.9410
|
| 47 |
+
Confusion Matrix:
|
| 48 |
+
[[2425 75]
|
| 49 |
+
[ 220 2280]]
|
| 50 |
+
|
| 51 |
+
---
|
| 52 |
+
|
| 53 |
+
You can download the model from [Hugging Face](https://huggingface.co/Wolverine001/Alexnet-TransferLearning).
|