๐ [1] to deal with folder of training Images <use python os,shutil>
๐ https://www.geeksforgeeks.org/os-module-python-examples/
๐ https://docs.python.org/3/library/shutil.html
๐ [2] to display images in run mode <use seaborn>
๐ https://seaborn.pydata.org/
๐ [3] to prepare Images to training <use torchvision transform>
๐ https://pytorch.org/vision/stable/generated/torchvision.transforms.Compose.html
๐ [4] to convert the image folder to dataset format <use torchvision dataset>
๐ https://pytorch.org/vision/stable/datasets.html
๐ [5] to split dataset to training and testing <use torch random_split>
๐ https://pytorch.org/docs/stable/data.html#torch.utils.data.random_split
๐ [6] to combine a dataset(train & test) and a sampler <use torch DataLoader>
๐ https://pytorch.org/docs/stable/data.html#torch.utils.data.DataLoader
๐ [7] to show image in grid format <use torchvision make_grid>
๐ https://pytorch.org/vision/stable/generated/torchvision.utils.make_grid.html
๐ [8] before creating model you should using GPU (CUDA) device to create a model
๐ https://pytorch.org/docs/stable/cuda.html
๐ [9] to normalized the data to make model run better <use torchvision transforms>
๐ https://pytorch.org/vision/stable/generated/torchvision.transforms.Compose.html
๐ [10] to create the model we will use the (resnet50) model
๐ https://pytorch.org/vision/main/models/generated/torchvision.models.resnet50.html
๐ https://pytorch.org/docs/stable/generated/torch.nn.Linear.html
๐ https://pytorch.org/docs/stable/generated/torch.nn.CrossEntropyLoss.html
๐ https://pytorch.org/docs/stable/generated/torch.optim.SGD.html
๐ https://pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html
๐ https://pytorch.org/docs/stable/generated/torch.no_grad.html
๐ [11] to save the model after training <use torch.save>
๐ https://pytorch.org/docs/stable/generated/torch.save.html
๐ [12] to test the model using Images
๐ https://pytorch.org/vision/stable/generated/torchvision.datasets.ImageFolder.html
๐ normalize image go to step number [9]
๐ https://pytorch.org/docs/stable/generated/torch.unsqueeze.html
๐ https://pytorch.org/docs/stable/generated/torch.nn.Softmax.html