Pytorch-Tensor-Basics.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Pytorch-Tensor-Basics.ipynb","provenance":[],"collapsed_sections":[]},"kernelspec":{"name":"python3","display_name":"Python 3"},"accelerator":"GPU"},"cells":[{"cell_type":"code","metadata":{"id":"vYnr1MzOdiDe","colab_type":"code","outputId":"894b8537-78db-490e-c663-72700afe5908","executionInfo":{"status":"ok","timestamp":1571365991380,"user_tz":240,"elapsed":2001,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":34}},"source":["import torch\n","\n","print(torch.__version__)"],"execution_count":2,"outputs":[{"output_type":"stream","text":["1.2.0\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"IESwlsjOfZvN","colab_type":"text"},"source":["**1. Define Tensor**\n","\n","Tensor is a data structure used in PyTorch. \n","\n","Tensors are similar to NumPy’s ndarrays, with the addition being that Tensors can also be used on a GPU to accelerate computing.\n","\n","![alt text](https://www.kdnuggets.com/wp-content/uploads/tensor-examples.jpg)"]},{"cell_type":"code","metadata":{"id":"qF1wg0Hvd2lS","colab_type":"code","outputId":"0fb623f6-a3d1-4ce4-d291-9de476c17d1b","executionInfo":{"status":"ok","timestamp":1571160674677,"user_tz":240,"elapsed":257,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":121}},"source":["x = torch.empty(5, 3)\n","print(x)\n","print(x.shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[2.5636e-36, 0.0000e+00, 4.4842e-44],\n","        [0.0000e+00,        nan, 5.6052e-44],\n","        [8.4034e+20, 3.3459e+21, 5.3697e-05],\n","        [1.3721e-05, 5.4649e-05, 6.7651e+22],\n","        [6.4098e-10, 1.6765e+22, 4.6291e-01]])\n","torch.Size([5, 3])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"LA-Nj81PeJXA","colab_type":"code","outputId":"3cf40852-8b90-455d-849e-a033bf1116e4","executionInfo":{"status":"ok","timestamp":1571160642101,"user_tz":240,"elapsed":306,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":104}},"source":["x = torch.rand(5, 3)\n","print(x)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.8374, 0.6568, 0.9469],\n","        [0.5035, 0.1892, 0.6594],\n","        [0.1111, 0.9556, 0.1687],\n","        [0.3096, 0.7158, 0.5957],\n","        [0.0495, 0.7191, 0.4629]])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"lpi0VYf-DQ9S","colab_type":"code","outputId":"04c9f141-a8ed-40da-8c69-598d4b22c2f9","executionInfo":{"status":"ok","timestamp":1571338448722,"user_tz":240,"elapsed":1182,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":191}},"source":["import numpy as np\n","\n","x = np.ones((5, 3))\n","y = torch.from_numpy(x)\n","\n","print(x)\n","print(y)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["[[1. 1. 1.]\n"," [1. 1. 1.]\n"," [1. 1. 1.]\n"," [1. 1. 1.]\n"," [1. 1. 1.]]\n","tensor([[1., 1., 1.],\n","        [1., 1., 1.],\n","        [1., 1., 1.],\n","        [1., 1., 1.],\n","        [1., 1., 1.]], dtype=torch.float64)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"bmZeCTtleXT4","colab_type":"code","outputId":"2e565ae7-57ec-487e-84ec-14429654d483","executionInfo":{"status":"ok","timestamp":1571160759217,"user_tz":240,"elapsed":268,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":104}},"source":["x = torch.Tensor([[2,1,2],[2,3,6],[3,4,8],[9,1,2],[4,5,2]])\n","print(x)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[2., 1., 2.],\n","        [2., 3., 6.],\n","        [3., 4., 8.],\n","        [9., 1., 2.],\n","        [4., 5., 2.]])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"XuxBLRwKt0qn","colab_type":"code","outputId":"7d001d1b-0854-4f82-f046-f0bf717a1799","executionInfo":{"status":"ok","timestamp":1571164781596,"user_tz":240,"elapsed":233,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":34}},"source":["x = torch.ones(4, 5, 6, 7)\n","print(x.shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["torch.Size([4, 5, 6, 7])\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"BHi_KfP2e9wb","colab_type":"text"},"source":["**2. Index Tensor**"]},{"cell_type":"code","metadata":{"id":"-_ECtUthfD6N","colab_type":"code","outputId":"afcd8298-2bf8-42fd-f3d2-af56435d2f01","executionInfo":{"status":"ok","timestamp":1571164869496,"user_tz":240,"elapsed":338,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":139}},"source":["x = torch.rand(2,3,2)\n","print(x)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[[0.7446, 0.9678],\n","         [0.9088, 0.7623],\n","         [0.9062, 0.6040]],\n","\n","        [[0.5469, 0.4662],\n","         [0.5433, 0.2112],\n","         [0.2959, 0.8491]]])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"Pv3eNeGFuWLq","colab_type":"code","outputId":"090ba22e-cabc-4cf9-d4e7-247ab6606381","executionInfo":{"status":"ok","timestamp":1571165068062,"user_tz":240,"elapsed":264,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":86}},"source":["print(x[1])\n","print(x[1].shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.5469, 0.4662],\n","        [0.5433, 0.2112],\n","        [0.2959, 0.8491]])\n","torch.Size([3, 2])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"7goxMMpLvGFn","colab_type":"code","outputId":"e6ee1e10-09a9-4ea1-a2f0-ffb1e71fcf33","executionInfo":{"status":"ok","timestamp":1571165076269,"user_tz":240,"elapsed":270,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":86}},"source":["print(x[-1])\n","print(x[-1].shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.5469, 0.4662],\n","        [0.5433, 0.2112],\n","        [0.2959, 0.8491]])\n","torch.Size([3, 2])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"pIusjD3AuZjS","colab_type":"code","outputId":"bc87586f-ff83-4681-b075-e2d505f199a9","executionInfo":{"status":"ok","timestamp":1571165020536,"user_tz":240,"elapsed":289,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":69}},"source":["print(x[:,0])\n","print(x[:,0].shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.7446, 0.9678],\n","        [0.5469, 0.4662]])\n","torch.Size([2, 2])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"E0o4Uax2u70d","colab_type":"code","outputId":"ece05392-446c-4407-cec2-4965a9cd5e5b","executionInfo":{"status":"ok","timestamp":1571165037377,"user_tz":240,"elapsed":477,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":121}},"source":["print(x[:,1:3])\n","print(x[:,1:3].shape)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[[0.9088, 0.7623],\n","         [0.9062, 0.6040]],\n","\n","        [[0.5433, 0.2112],\n","         [0.2959, 0.8491]]])\n","torch.Size([2, 2, 2])\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"hYpAyyzatKS6","colab_type":"text"},"source":["**3. Reshape Tensor**"]},{"cell_type":"code","metadata":{"id":"OAmm2tV2tSd1","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":104},"outputId":"7dfc8e41-e5e2-49b5-cc44-e4112c93d6cf","executionInfo":{"status":"ok","timestamp":1571366679276,"user_tz":240,"elapsed":494,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}}},"source":["x = torch.rand(4,3,2)\n","print(x.shape)\n","\n","\n","print(x.view(12, 2).shape) # .view(): Returns a new tensor with the same data as the self tensor but of a different shape.\n","print(x.view(-1, 2).shape) # If one of the dimensions is -1, its size can be inferred\n","print(x.view(4, -1).shape)\n","print(x.view(-1).shape)"],"execution_count":25,"outputs":[{"output_type":"stream","text":["torch.Size([4, 3, 2])\n","torch.Size([12, 2])\n","torch.Size([12, 2])\n","torch.Size([4, 6])\n","torch.Size([24])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"MUL8ojJzts8x","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":52},"outputId":"5de75550-f99b-4aab-b6e8-94cec8310723","executionInfo":{"status":"ok","timestamp":1571366680756,"user_tz":240,"elapsed":468,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}}},"source":["print(x.permute([1,0,2]).shape)\n","print(x.permute([2,1,0]).shape)"],"execution_count":26,"outputs":[{"output_type":"stream","text":["torch.Size([3, 4, 2])\n","torch.Size([2, 3, 4])\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"rLQJSYy2t939","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":69},"outputId":"fa128401-dfc4-437d-8eb6-a2c85fafe438","executionInfo":{"status":"ok","timestamp":1571366719485,"user_tz":240,"elapsed":1167,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}}},"source":["print(x.unsqueeze(0).shape)           # .unsqueeze(): Returns a new tensor with a dimension of size one inserted at the specified position.\n","print(x.unsqueeze(-1).shape)\n","\n","unsqueezed_x = x.unsqueeze(-1)\n","print(unsqueezed_x.squeeze().shape)   # .squeeze(): Returns a tensor with all the dimensions of input of size 1 removed."],"execution_count":28,"outputs":[{"output_type":"stream","text":["torch.Size([1, 4, 3, 2])\n","torch.Size([4, 3, 2, 1])\n","torch.Size([4, 3, 2])\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"GFCTtiKdUNkl","colab_type":"text"},"source":["**4. CUDA Extensions**\n","\n","![alt text](https://2.bp.blogspot.com/-hNBj-v6c5As/WR8UefJAFkI/AAAAAAAAAis/2KXxy8EgSSE6gsLoj6dbuQZaYnWdofrcACLcB/s400/GPU_vs_CPU.png)\n","\n","![alt text](https://4.bp.blogspot.com/-3I8tFgCq4Xg/WSaoo8sI9KI/AAAAAAAAAjc/wA06nUNFFus6t5tpoVmWoBXrQXwtdLQbgCLcB/s640/pytorch_cpu_v_gpu_2.png)"]},{"cell_type":"code","metadata":{"id":"btc2c-6-UNKN","colab_type":"code","outputId":"651cb5fc-a5b2-4e47-9a17-ef06c5e07b3e","executionInfo":{"status":"ok","timestamp":1571174979609,"user_tz":240,"elapsed":454,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":52}},"source":["print(torch.cuda.is_available())\n","print(torch.cuda.device_count())"],"execution_count":0,"outputs":[{"output_type":"stream","text":["True\n","1\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"UY74nR4lUXHK","colab_type":"code","outputId":"773d1e58-e29b-474d-bf14-248620f202a4","executionInfo":{"status":"ok","timestamp":1571325557561,"user_tz":240,"elapsed":1220,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":312}},"source":["!nvidia-smi"],"execution_count":0,"outputs":[{"output_type":"stream","text":["Thu Oct 17 15:19:16 2019       \n","+-----------------------------------------------------------------------------+\n","| NVIDIA-SMI 430.40       Driver Version: 418.67       CUDA Version: 10.1     |\n","|-------------------------------+----------------------+----------------------+\n","| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n","| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n","|===============================+======================+======================|\n","|   0  Tesla K80           Off  | 00000000:00:04.0 Off |                    0 |\n","| N/A   46C    P0    60W / 149W |      0MiB / 11441MiB |      0%      Default |\n","+-------------------------------+----------------------+----------------------+\n","                                                                               \n","+-----------------------------------------------------------------------------+\n","| Processes:                                                       GPU Memory |\n","|  GPU       PID   Type   Process name                             Usage      |\n","|=============================================================================|\n","|  No running processes found                                                 |\n","+-----------------------------------------------------------------------------+\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"VF7gqtpGVDOb","colab_type":"code","outputId":"f663ac00-6bfe-416b-a9ff-288afa1132c0","executionInfo":{"status":"ok","timestamp":1571175025587,"user_tz":240,"elapsed":316,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":121}},"source":["x = torch.rand(5,3)\n","print(x)\n","print(x.device)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.7391, 0.0859, 0.0604],\n","        [0.3890, 0.7486, 0.5022],\n","        [0.8036, 0.9152, 0.9488],\n","        [0.3435, 0.7254, 0.2349],\n","        [0.5282, 0.5450, 0.1722]])\n","cpu\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"gnZTuHbqUjI_","colab_type":"code","outputId":"bb8e6a5f-4562-48f5-8324-361aefbb66e6","executionInfo":{"status":"ok","timestamp":1571175034682,"user_tz":240,"elapsed":328,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":121}},"source":["device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n","\n","x = x.to(device)\n","print(x)\n","print(x.device)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["tensor([[0.7391, 0.0859, 0.0604],\n","        [0.3890, 0.7486, 0.5022],\n","        [0.8036, 0.9152, 0.9488],\n","        [0.3435, 0.7254, 0.2349],\n","        [0.5282, 0.5450, 0.1722]], device='cuda:0')\n","cuda:0\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"Z5DE7kgoVODs","colab_type":"code","outputId":"e692b02c-5ef6-4b7f-df93-08f1123f3e60","executionInfo":{"status":"ok","timestamp":1571175192877,"user_tz":240,"elapsed":334,"user":{"displayName":"Minju Jung","photoUrl":"","userId":"09046723808353533857"}},"colab":{"base_uri":"https://localhost:8080/","height":69}},"source":["x = torch.rand(5,3)\n","print(x.device)\n","\n","x = x.cuda()\n","print(x.device)\n","\n","x = x.cpu()\n","print(x.device)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["cpu\n","cuda:0\n","cpu\n"],"name":"stdout"}]}]}