{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import torch\nimport torchvision.datasets as data\nimport torchvision.transforms as transforms\nimport random","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","collapsed":true,"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":false},"cell_type":"code","source":"device = 'cuda' if torch.cuda.is_available() else 'cpu'\n\nrandom.seed(10024)\ntorch.manual_seed(10024)\nif device == 'cuda':\n  torch.cuda.manual_seed_all(10024)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\n# 학습 파라미터 설정\nlearning_rate = 0.005\ntraining_epochs = 15\nbatch_size = 100","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!pip uninstall -y kaggle\n!pip install --upgrade pip\n!pip install kaggle==1.5.6\n!mkdir -p ~/.kaggle\n!cp kaggle.json ~/.kaggle/\n!chmod 600 ~/.kaggle/kaggle.json","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!kaggle competitions download -c 2020-ai-exam-fashionmnist-3\n!unzip 2020-ai-exam-fashionmnist-3.zip","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import pandas as pd\nimport numpy as np","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data=pd.read_csv('mnist_train_label.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data=pd.read_csv('mnist_train_label.csv',header=None, usecols=range(0,785))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn import preprocessing\nScaler = preprocessing.StandardScaler()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"xy_train_data=train_data.to_numpy()\n\n\nx_train_data=xy_train_data[:,1:]\nx_train_data = Scaler.fit_transform(x_train_data)\ny_train_data=xy_train_data[:,0]\n\nx_train_data=torch.FloatTensor(x_train_data)\ny_train_data=torch.LongTensor(y_train_data)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_train_data.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_dataset = torch.utils.data.TensorDataset(x_train_data, y_train_data)\ndata_loader = torch.utils.data.DataLoader(dataset=train_dataset,\n                                          batch_size=batch_size,\n                                          shuffle=True,\n                                          drop_last=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"linear1 = torch.nn.Linear(784,512,bias=True)\nlinear2 = torch.nn.Linear(512,512,bias=True)\nlinear3 = torch.nn.Linear(512,10,bias=True)\nrelu = torch.nn.ReLU()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"torch.nn.init.normal_(linear1.weight)\ntorch.nn.init.normal_(linear2.weight)\ntorch.nn.init.normal_(linear3.weight)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# ======================================\n# relu는 맨 마지막 레이어에서 빼는 것이 좋다.\n# ======================================\n# model = torch.nn.Sequential(linear1,relu,linear2,relu,linear3,relu).to(device) # 주의사항\nmodel = torch.nn.Sequential(linear1,relu,\n                            linear2,relu,\n                            linear3).to(device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# 손실함수와 최적화 함수\nloss = torch.nn.CrossEntropyLoss().to(device) # softmax 내부적으로 계산\noptimizer = torch.optim.Adam(model.parameters(), lr=learning_rate) ","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\ntotal_batch = len(data_loader)\nfor epoch in range(training_epochs):\n    avg_cost = 0\n\n    for X, Y in data_loader:\n\n        # (1000, 1, 28, 28) 크기의 텐서를 (1000, 784) 크기의 텐서로 변형\n        X = X.to(device)\n        # one-hot encoding되어 있지 않음\n        Y = Y.to(device)\n        #%debug\n\n        # 그래디언트 초기화\n        optimizer.zero_grad()\n        # Forward 계산\n        hypothesis = model(X)\n        # Error 계산\n        cost = loss(hypothesis, Y)\n        # Backparopagation\n        cost.backward()\n        # 가중치 갱신\n        optimizer.step()\n\n        # 평균 Error 계산\n        avg_cost += cost / total_batch\n\n    print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost))\n\nprint('Learning finished')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test = pd.read_csv('mnist_test.csv',header=None, usecols=range(0,784))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test = x_test.to_numpy()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test[:,0] = 0","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_test = Scaler.transform(x_test)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Test the model using test sets\nwith torch.no_grad():\n\n  \n  x_test_data=torch.from_numpy(x_test).float().to(device)\n\n  prediction = model(x_test_data)\n  correct_prediction = torch.argmax(prediction, 1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"correct_prediction = correct_prediction.cpu().numpy().reshape(-1,1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"correct_prediction","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submit=pd.read_csv('submission.csv')\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"for i in range(len(correct_prediction)):\n  submit['Category'][i]=correct_prediction[i].item()\nsubmit.to_csv('submit.csv',index=False,header=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\n!kaggle competitions submit -c 2020-ai-exam-fashionmnist-3 -f submit.csv -m \"submit\"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}