File size: 2,270 Bytes
23552e8 b51c3f7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 |
---
tags:
- model_hub_mixin
- pytorch_model_hub_mixin
---
# Model Card: MRI Brain Tumor Classification (ResNet-18)
## Model Details
- **Model Name**: `MRIResnet`
- **Architecture**: ResNet-18-based model for MRI brain tumor classification
- **Dataset**: [Brain Tumor MRI Dataset](https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset)
- **Batch Size**: 32
- **Loss Function**: CrossEntropy Loss
- **Optimizer**: Adam (learning rate = 1e-3)
- **Transfer Learning**: Yes (pretrained ResNet-18 with modified layers)
## Model Architecture
This model is based on **ResNet-18**, a widely used convolutional neural network, and has been adapted for **MRI-based brain tumor classification**.
### **Modifications**
- **Input Channel Adaptation**: The first convolutional layer is modified to accept single-channel (grayscale) MRI scans.
- **Classifier Head**: The fully connected (FC) layer is replaced to output 4 classes (assuming 4 tumor categories).
- **Transfer Learning**:
- **Frozen Layers**: All pre-trained weights are frozen except for the modified layers.
- **Trainable Layers**:
- First convolutional layer (`conv1`)
- Fully connected classification layer (`fc`)
## Implementation
### **Model Definition**
```python
import torch
import torch.nn as nn
from torchvision.models import resnet18
class MRIResnet(nn.Module, PyTorchModelHubMixin):
def __init__(self):
super().__init__()
self.base_model = resnet18(weights=True)
self.base_model.conv1 = nn.Conv2d(
1, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False
)
self.base_model.fc = nn.Linear(512, 4)
# Freeze all layers except the modified ones
for param in self.base_model.parameters():
param.requires_grad = False
for param in self.base_model.conv1.parameters():
param.requires_grad = True
for param in self.base_model.fc.parameters():
param.requires_grad = True
def forward(self, x):
return self.base_model(x)
```
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration: |