Clone this repository and install the dependencies:
git clone https://github.com/liangchingyun/A-Convolution-Module-For-Variable-Input-Channels.git
cd A-Convolution-Module-For-Variable-Input-Channels
pip install -r requirements.txt
Make sure you have Jupyter Notebook installed. If not, you can install it using the following command:
pip install jupyter
Navigate to the cloned repository directory and open the main.ipynb file.
You can modify the training parameters below to customize your training process. Here's a brief description of each parameter:
batch_size
: The batch size used for training.test_batch_size
: The batch size used for testing.epochs
: The number of epochs to train the model.learning_rate
: The learning rate used by the optimizer.momentum
: The momentum used by the optimizer.weight_decay
: The weight decay (L2 penalty) applied to the model parameters.
To change these parameters, simply update their values in the code cell below.
parser.add_argument('--batch-size', type=int, default=256)
parser.add_argument('--test-batch-size', type=int, default=20)
parser.add_argument('--epochs', type=int, default=100)
parser.add_argument('--lr', type=float, default=0.1, )
parser.add_argument('--momentum', type=float, default=0.9)
parser.add_argument('--weight-decay', type=float, default=1e-4)
parser.add_argument('--net-name', default='resnet18')
parser.add_argument('--channel', default='RG')