sebastiansarasti
commited on
Commit
•
7e4cbe5
1
Parent(s):
72a07e3
Update README.md
Browse files
README.md
CHANGED
@@ -1,9 +1,55 @@
|
|
1 |
---
|
2 |
tags:
|
3 |
-
-
|
4 |
-
- pytorch_model_hub_mixin
|
5 |
---
|
6 |
|
7 |
-
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
tags:
|
3 |
+
- pytorch
|
|
|
4 |
---
|
5 |
|
6 |
+
# CNN Leukemia Classifier
|
7 |
+
|
8 |
+
## Model Description
|
9 |
+
|
10 |
+
This Convolutional Neural Network (CNN) model is designed for the classification of leukemia images into one of four classes. It was developed for the Quito AI Day event by [Sebastian Sarasti](https://www.linkedin.com/in/sebastiansarasti/). The model leverages a series of convolutional layers followed by fully connected layers to process and classify images effectively.
|
11 |
+
|
12 |
+
## Model Architecture
|
13 |
+
|
14 |
+
The model consists of the following layers:
|
15 |
+
|
16 |
+
- Convolutional Layer: 3 input channels, 128 output channels, 3x3 kernel size, stride 1, padding 1
|
17 |
+
- ReLU Activation
|
18 |
+
- Max Pooling Layer: 2x2 kernel size
|
19 |
+
- Dropout Layer: 0.3 dropout rate
|
20 |
+
|
21 |
+
- Convolutional Layer: 128 input channels, 64 output channels, 3x3 kernel size, stride 1, padding 1
|
22 |
+
- ReLU Activation
|
23 |
+
- Max Pooling Layer: 2x2 kernel size
|
24 |
+
- Dropout Layer: 0.3 dropout rate
|
25 |
+
|
26 |
+
- Convolutional Layer: 64 input channels, 32 output channels, 3x3 kernel size, stride 1, padding 1
|
27 |
+
- ReLU Activation
|
28 |
+
- Max Pooling Layer: 2x2 kernel size
|
29 |
+
- Dropout Layer: 0.3 dropout rate
|
30 |
+
|
31 |
+
- Convolutional Layer: 32 input channels, 8 output channels, 3x3 kernel size, stride 1, padding 1
|
32 |
+
- ReLU Activation
|
33 |
+
- Max Pooling Layer: 2x2 kernel size
|
34 |
+
- Dropout Layer: 0.3 dropout rate
|
35 |
+
|
36 |
+
- Flatten Layer
|
37 |
+
|
38 |
+
- Fully Connected Layer: 1568 input features, 512 output features
|
39 |
+
- ReLU Activation
|
40 |
+
- Dropout Layer: 0.5 dropout rate
|
41 |
+
|
42 |
+
- Fully Connected Layer: 512 input features, 4 output features
|
43 |
+
|
44 |
+
## Dataset
|
45 |
+
|
46 |
+
The model was trained on the [Leukemia dataset from Kaggle](https://www.kaggle.com/datasets/mehradaria/leukemia), which consists of images labeled into different leukemia types.
|
47 |
+
|
48 |
+
## Usage
|
49 |
+
|
50 |
+
To use this model, you can load it from the Hugging Face Hub as follows:
|
51 |
+
|
52 |
+
```python
|
53 |
+
from transformers import AutoModel
|
54 |
+
|
55 |
+
model = AutoModel.from_pretrained("path/to/your/model")
|