{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "0oTJizeDuDnC" }, "source": [ "# CSC401/485 F24 Tutorial - Introduction to Pytorch\n", "\n", "By Gavin Guan. Based on tutorial created by Frank Niu, Zhewei Sun." ] }, { "cell_type": "markdown", "metadata": { "id": "Xq2d6RbQu8Yh" }, "source": [ "## About Pytorch\n", "\n", "We will be using PyTorch 2.3.0 for this assignment.\n", "This version has been pre-installed on teach.cs with Python 3.12.\n", "To invoke Python 3.12, use the command `python3.12` instead of `python` when running your code.\n", "\n", "\n", "**Useful Links:**\n", "\n", "- PyTorch documentation: https://pytorch.org/docs/2.3/\n", "- Eionps: https://einops.rocks/" ] }, { "cell_type": "code", "execution_count": 126, "metadata": { "id": "x6H16q_SwKIu" }, "outputs": [], "source": [ "import torch" ] }, { "cell_type": "markdown", "metadata": { "id": "evnjdqNehFVy" }, "source": [ "### Overview\n", "\n", "In essense, PyTorch is a Python based deep learning package that allows you to specify, train, and execute your own neural networks efficiently.\n", "\n", "A complete PyTorch program typically involve the following steps:\n", "\n", "0. Prepare your data and specify your neural network.\n", "1. Load a (batch of) datapoint(s) (i.e. dataloader).\n", "2. Pass the datapoint(s) into your neural network (i.e. forward pass).\n", "3. Compute loss based on the neural network's output.\n", "4. Back-propagate error and adjust weights of the neural network (i.e. backward pass).\n", "\n", "A package like PyTorch trivializes back-prop with a powerful auto-differentiation engine.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "tqkYUUo-hKpd" }, "source": [ "## 1. Working with PyTorch Tensors" ] }, { "cell_type": "markdown", "metadata": { "id": "YEJ8NbcahUp1" }, "source": [ "### 1.1 Creating Tensors\n", "\n", "A **tensor** is basically a PyTorch object that stores your data:" ] }, { "cell_type": "code", "execution_count": 127, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "F3Vw5tImhaRc", "outputId": "a960c927-3af9-4c0c-b45e-561d2848f96b" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]])" ] }, "execution_count": 127, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor = torch.tensor([[1,2], [3,4], [5,6]], dtype=torch.float32)\n", "my_tensor" ] }, { "cell_type": "markdown", "metadata": { "id": "EEHRI_n0hrmd" }, "source": [ "Similar to numpy, you can create pre-populated tensors:\n" ] }, { "cell_type": "code", "execution_count": 128, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "8SywnC0jhshd", "outputId": "0dff43b6-7e07-47b5-e27d-7354ed2ed00c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[1., 1., 1., 1.],\n", " [1., 1., 1., 1.],\n", " [1., 1., 1., 1.]])\n", "tensor([[0., 0., 0., 0.],\n", " [0., 0., 0., 0.]])\n", "tensor([[0, 1, 2],\n", " [3, 4, 5]])\n" ] } ], "source": [ "print(torch.ones((3,4)))\n", "print(torch.zeros((2,4)))\n", "print(torch.arange(6).view(2, 3))" ] }, { "cell_type": "markdown", "metadata": { "id": "hYOsUspJ4CZk" }, "source": [ "### 1.2 Tensor Attributes" ] }, { "cell_type": "code", "execution_count": 129, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "cNfHELz-4CZk", "outputId": "107d7e8b-c82a-4c5d-e57a-31e871a53085" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]])" ] }, "execution_count": 129, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor" ] }, { "cell_type": "markdown", "metadata": { "id": "9lMUDT414CZl" }, "source": [ "**.size()** or **.shape** returns the size of the tensor:" ] }, { "cell_type": "code", "execution_count": 131, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "DYOmvcje4CZl", "outputId": "923472b2-8414-481e-e9e9-a9fded04c171" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "my_tensor.size()=torch.Size([3, 2])\n", "my_tensor.shape=torch.Size([3, 2])\n", "my_tensor.size()[0]=3\n", "my_tensor.size(1)=2\n" ] } ], "source": [ "print(f\"{my_tensor.size()=}\")\n", "print(f\"{my_tensor.shape=}\")\n", "print(f\"{my_tensor.size()[0]=}\")\n", "print(f\"{my_tensor.size(1)=}\")" ] }, { "cell_type": "markdown", "metadata": { "id": "w1I50gdC4CZm" }, "source": [ "**.dim()** returns the number of dimensions your tensor has\n", "\n", "**.dtype** tells you the type of the data stored within the tensor" ] }, { "cell_type": "code", "execution_count": 133, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "SSGNrBp34CZm", "outputId": "e80db903-1711-48f3-8480-642c828dca5c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "my_tensor.dim()=2\n", "my_tensor.dtype=torch.float32\n" ] } ], "source": [ "print(f\"{my_tensor.dim()=}\")\n", "print(f\"{my_tensor.dtype=}\")" ] }, { "cell_type": "markdown", "metadata": { "id": "ZNKnN9UE4CZn" }, "source": [ "**.requires_grad** tells you whether the tensor is expected receive back-propagated gradient. You can set this to true by specifying *requires_grad=True* in the constructor." ] }, { "cell_type": "code", "execution_count": 134, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "rrt2K4ys4CZn", "outputId": "a0bcf22a-3a35-4a3f-9ae9-cc4b82994395" }, "outputs": [ { "data": { "text/plain": [ "False" ] }, "execution_count": 134, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor.requires_grad" ] }, { "cell_type": "code", "execution_count": 135, "metadata": { "id": "kYjSeJi-4CZn" }, "outputs": [], "source": [ "my_tensor = torch.tensor([[1,2], [3,4], [5,6]], dtype=torch.float32, requires_grad=True)" ] }, { "cell_type": "code", "execution_count": 136, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "kDkYVLRv4CZn", "outputId": "4d50a38d-0429-4cc6-82a1-74858acf0999" }, "outputs": [ { "data": { "text/plain": [ "True" ] }, "execution_count": 136, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor.requires_grad" ] }, { "cell_type": "code", "execution_count": 137, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "1hkAbCRk4CZo", "outputId": "e1555b65-b5ef-479f-be50-333ad40cdc71" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], requires_grad=True)" ] }, "execution_count": 137, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor" ] }, { "cell_type": "markdown", "metadata": { "id": "y_LrMEoO4CZo" }, "source": [ "**.device** tells you which device memory the tensor sits in. This can also be specified when constructing the tensor." ] }, { "cell_type": "code", "execution_count": 141, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "dxABW_KZ4CZo", "outputId": "7ee649b2-ad3e-418c-af4e-05f81ab42866" }, "outputs": [ { "data": { "text/plain": [ "device(type='cuda', index=0)" ] }, "execution_count": 141, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor.device" ] }, { "cell_type": "code", "execution_count": 139, "metadata": { "id": "VrCZQd8r4CZo" }, "outputs": [], "source": [ "#device = torch.device('cpu')\n", "### Try this if you have a gpu.\n", "### The number after the colon specifies which gpu to use (if you're fortunate enough to have more than 1, that is).\n", "device = torch.device('cuda:0')" ] }, { "cell_type": "code", "execution_count": 140, "metadata": { "id": "qSUiOfnX4CZo" }, "outputs": [], "source": [ "my_tensor = torch.tensor([[1,2], [3,4], [5,6]], dtype=torch.float32, requires_grad=True, device=device)" ] }, { "cell_type": "markdown", "metadata": { "id": "yqCKdENG4CZp" }, "source": [ "Use **.to** to explicitly send a tensor to a specific device:" ] }, { "cell_type": "code", "execution_count": 142, "metadata": { "id": "hJ737pYN4CZp" }, "outputs": [], "source": [ "my_tensor = my_tensor.to(device)" ] }, { "cell_type": "markdown", "metadata": { "id": "t1Djh0x44CZp" }, "source": [ "### 1.3 Tensor Operations" ] }, { "cell_type": "code", "execution_count": 149, "metadata": { "id": "VPh9lYXc4CZp" }, "outputs": [], "source": [ "tensor_a = torch.tensor([[1,2], [3,4]], dtype=torch.float32, requires_grad=True)\n", "tensor_b = torch.tensor([[[.5, .6], [.7, .8]]], dtype=torch.float32, requires_grad=True)" ] }, { "cell_type": "markdown", "metadata": { "id": "8z7AocE34CZq" }, "source": [ "You can apply arithmetic operations to tensors just like numpy arrays:" ] }, { "cell_type": "code", "execution_count": 150, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "o3XZoi0h4CZq", "outputId": "92eb0e5f-5654-4af9-b45a-f86a371801cc" }, "outputs": [ { "data": { "text/plain": [ "tensor([[[1.5000, 2.6000],\n", " [3.7000, 4.8000]]], grad_fn=)" ] }, "execution_count": 150, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_a + tensor_b" ] }, { "cell_type": "markdown", "metadata": { "id": "AxD9J0_v4CZq" }, "source": [ "Notice how the tensor keeps track of extra information because we need the gradients to be back-propogated." ] }, { "cell_type": "code", "execution_count": 151, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "dU3WwsGI4CZq", "outputId": "349007a1-8f70-4139-b185-822687588f7d" }, "outputs": [ { "data": { "text/plain": [ "tensor([[2., 4.],\n", " [6., 8.]], grad_fn=)" ] }, "execution_count": 151, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_a * 2" ] }, { "cell_type": "markdown", "metadata": { "id": "qtaxq_Rf4CZr" }, "source": [ "Applying the * operator to two tensors does an element-wise multiplication:" ] }, { "cell_type": "code", "execution_count": 152, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "i75dcn4j4CZr", "outputId": "7a6acd9b-d99f-44d7-aff3-4dc81bbcc012", "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "tensor([[[0.5000, 1.2000],\n", " [2.1000, 3.2000]]], grad_fn=)" ] }, "execution_count": 152, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_a * tensor_b" ] }, { "cell_type": "markdown", "metadata": { "id": "XgLipvk34CZr" }, "source": [ "### 1.4 Tensor Functions" ] }, { "cell_type": "markdown", "metadata": { "id": "wBxL9nAp4CZs" }, "source": [ "Use **torch.matmul** for matrix multiplication:" ] }, { "cell_type": "code", "execution_count": 154, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Sa4jamIN4CZs", "outputId": "7abcb44c-19db-46a4-b08d-78a90ca65c3b", "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "tensor([[[1.9000, 2.2000],\n", " [4.3000, 5.0000]]], grad_fn=)" ] }, "execution_count": 154, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.matmul(tensor_a, tensor_b)" ] }, { "cell_type": "markdown", "metadata": { "id": "N9x1BpyEd900" }, "source": [ "Most of the tensor functions can be called either as a torch function `function(a, b)`, or a tensor method `a.function(b)`." ] }, { "cell_type": "code", "execution_count": 155, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "PWlmoiuueQSQ", "outputId": "3ac4e587-7000-4849-b2a3-6d32cdd8356b" }, "outputs": [ { "data": { "text/plain": [ "tensor([[[1.9000, 2.2000],\n", " [4.3000, 5.0000]]], grad_fn=)" ] }, "execution_count": 155, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_a.matmul(tensor_b)" ] }, { "cell_type": "markdown", "metadata": { "id": "PYv7kbp9eTql" }, "source": [ "`@` is a valid shorthand for **matmul**." ] }, { "cell_type": "code", "execution_count": 156, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "o7iXn5gEd6fz", "outputId": "907c518e-0728-451f-acf8-0167f23e05c6" }, "outputs": [ { "data": { "text/plain": [ "tensor([[[1.9000, 2.2000],\n", " [4.3000, 5.0000]]], grad_fn=)" ] }, "execution_count": 156, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_a @ tensor_b" ] }, { "cell_type": "markdown", "metadata": { "id": "xbLOiR804CZs" }, "source": [ "**torch.sum** sums up the tensor:" ] }, { "cell_type": "code", "execution_count": 157, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "ZbYI65Uwty4W", "outputId": "a7d9f8ca-4e64-4570-e0c0-cfaf394be77a" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', requires_grad=True)" ] }, "execution_count": 157, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor\n" ] }, { "cell_type": "code", "execution_count": 158, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "jgPm3w_w4CZt", "outputId": "2ea57cde-2409-449d-84cf-a41075a93e61" }, "outputs": [ { "data": { "text/plain": [ "tensor(21., device='cuda:0', grad_fn=)" ] }, "execution_count": 158, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.sum(my_tensor)" ] }, { "cell_type": "code", "execution_count": 159, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor(21., device='cuda:0', grad_fn=)" ] }, "execution_count": 159, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor.sum()" ] }, { "cell_type": "markdown", "metadata": { "id": "Xy622DBK4CZt" }, "source": [ "You can specify the *axis* parameter to only sum up a particular dimension:" ] }, { "cell_type": "code", "execution_count": 162, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "k-xl3VdF4CZt", "outputId": "007a83cf-736d-4d9a-d297-0f500b1d87e1" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.sum(my_tensor, dim=0)=tensor([ 9., 12.], device='cuda:0', grad_fn=)\n", "torch.sum(my_tensor, axis=1)=tensor([ 3., 7., 11.], device='cuda:0', grad_fn=)\n" ] } ], "source": [ "print(f\"{torch.sum(my_tensor, dim=0)=}\")\n", "print(f\"{torch.sum(my_tensor, axis=1)=}\")" ] }, { "cell_type": "code", "execution_count": 163, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "hTH5Lx0F4CZt", "outputId": "2e91f146-a75b-4b5c-b4d6-9df9a0fd79de" }, "outputs": [ { "data": { "text/plain": [ "tensor([ 3., 7., 11.], device='cuda:0', grad_fn=)" ] }, "execution_count": 163, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.sum(my_tensor, axis=1)" ] }, { "cell_type": "markdown", "metadata": { "id": "1ZzHjtuG4CZt" }, "source": [ "More basic functions:" ] }, { "cell_type": "code", "execution_count": 164, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "dC3xy5LMwA8M", "outputId": "d24d9a3b-01f2-4365-fb25-54881bdfd0c0" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', requires_grad=True)" ] }, "execution_count": 164, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor" ] }, { "cell_type": "code", "execution_count": 165, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "UIXJAVHJ4CZu", "outputId": "f1dc4525-e547-4725-f32d-0ebc68a93076" }, "outputs": [ { "data": { "text/plain": [ "tensor(1., device='cuda:0', grad_fn=)" ] }, "execution_count": 165, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.min(my_tensor)" ] }, { "cell_type": "code", "execution_count": 166, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "zubxJF1U4CZu", "outputId": "4b5c76ed-a732-49d1-d12e-fb669729535b" }, "outputs": [ { "data": { "text/plain": [ "torch.return_types.max(\n", "values=tensor([2., 4., 6.], device='cuda:0', grad_fn=),\n", "indices=tensor([1, 1, 1], device='cuda:0'))" ] }, "execution_count": 166, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.max(my_tensor, axis=1)" ] }, { "cell_type": "code", "execution_count": 167, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "GoEdiQUd4CZu", "outputId": "b5e5c1fe-9074-4a6e-88fa-5962782d091c" }, "outputs": [ { "data": { "text/plain": [ "tensor(0, device='cuda:0')" ] }, "execution_count": 167, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.argmin(my_tensor)" ] }, { "cell_type": "code", "execution_count": 168, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "F4aL74kj4CZu", "outputId": "ea689392-88b6-4e39-d2d1-d545793d60ea" }, "outputs": [ { "data": { "text/plain": [ "tensor([1, 1, 1], device='cuda:0')" ] }, "execution_count": 168, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.argmax(my_tensor, axis=1)" ] }, { "cell_type": "markdown", "metadata": { "id": "u7WE6v8X4CZv" }, "source": [ "Note how gradient is not tracked after applying the two non-differentiable functions above." ] }, { "cell_type": "code", "execution_count": 169, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "KBBT9CKH4CZv", "outputId": "b15ad35b-adbe-4c8d-e599-0f5410940623" }, "outputs": [ { "data": { "text/plain": [ "tensor([3., 4.], device='cuda:0', grad_fn=)" ] }, "execution_count": 169, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.mean(my_tensor, axis=0)" ] }, { "cell_type": "code", "execution_count": 170, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "3LQLAm6B4CZv", "outputId": "641c0ae3-a614-488e-e0a7-3af0eeb117d4" }, "outputs": [ { "data": { "text/plain": [ "tensor(1.8708, device='cuda:0', grad_fn=)" ] }, "execution_count": 170, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.std(my_tensor)" ] }, { "cell_type": "markdown", "metadata": { "id": "vFELD0pf4CZv" }, "source": [ "**torch.topk** returns only the top-k elements in the specifed dimension:" ] }, { "cell_type": "code", "execution_count": 172, "metadata": { "id": "h4_9vs434CZw" }, "outputs": [], "source": [ "t = torch.tensor([[100,2,3], [4,5,6], [7,8,9]])" ] }, { "cell_type": "code", "execution_count": 173, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Qfh7sQWE4CZw", "outputId": "7bc12748-75a4-4726-bf14-589ee7255fe5" }, "outputs": [ { "data": { "text/plain": [ "tensor([[100, 2, 3],\n", " [ 4, 5, 6],\n", " [ 7, 8, 9]])" ] }, "execution_count": 173, "metadata": {}, "output_type": "execute_result" } ], "source": [ "t" ] }, { "cell_type": "markdown", "metadata": { "id": "ecbsItmZ4CZw" }, "source": [ "For each row, grab the top 2 elements in each column:" ] }, { "cell_type": "code", "execution_count": 174, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "fwr9o8gA4CZw", "outputId": "605cae2f-0676-4b5b-d55a-eee1ace4655b" }, "outputs": [], "source": [ "a, _ = torch.topk(t, k=2, dim=1)" ] }, { "cell_type": "markdown", "metadata": { "id": "6yNeZG5m4CZw" }, "source": [ "For each column, grab the top 2 elements in each row:" ] }, { "cell_type": "code", "execution_count": 176, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "WVy08K0H4CZw", "outputId": "135a4679-51a8-4156-df91-d75636bfd9e6" }, "outputs": [ { "data": { "text/plain": [ "torch.return_types.topk(\n", "values=tensor([[100, 8, 9],\n", " [ 7, 5, 6]]),\n", "indices=tensor([[0, 2, 2],\n", " [2, 1, 1]]))" ] }, "execution_count": 176, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.topk(t, k=2, dim=0)" ] }, { "cell_type": "markdown", "metadata": { "id": "rEwW_c1V4CZx" }, "source": [ "### 1.5 Slicing and Indexing" ] }, { "cell_type": "markdown", "metadata": { "id": "sbL1dHLf4CZx" }, "source": [ "A few ways to slice/index a PyTorch tensor:" ] }, { "cell_type": "code", "execution_count": 177, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "X_vRVihu4CZx", "outputId": "17b6f75f-a9c8-4830-9c69-296f542093a0" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', requires_grad=True)" ] }, "execution_count": 177, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor" ] }, { "cell_type": "markdown", "metadata": { "id": "IIJOtWP24CZx" }, "source": [ "Grab the first row:" ] }, { "cell_type": "code", "execution_count": 178, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "9J-oLYTM4CZx", "outputId": "42a0d516-bc9e-450c-d548-c2144244f034" }, "outputs": [ { "data": { "text/plain": [ "tensor([1., 2.], device='cuda:0', grad_fn=)" ] }, "execution_count": 178, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor[0, :]" ] }, { "cell_type": "markdown", "metadata": { "id": "H8em551s4CZx" }, "source": [ "Get the second element of the last two rows:" ] }, { "cell_type": "code", "execution_count": 179, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "IByAtbC14CZy", "outputId": "f368ef5b-d60c-4974-c98e-55471a0d053f" }, "outputs": [ { "data": { "text/plain": [ "tensor([4., 6.], device='cuda:0', grad_fn=)" ] }, "execution_count": 179, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor[-2:, 1]" ] }, { "cell_type": "markdown", "metadata": { "id": "lDw1PSd14CZz" }, "source": [ "You can also use a integer tensor to index another tensor:" ] }, { "cell_type": "code", "execution_count": 184, "metadata": { "id": "8G2Y3GlT4CZz" }, "outputs": [], "source": [ "a = torch.tensor([0.0, 1.1, 2.2, 3.3, 4.4, 5.5, 2.2])\n", "b = torch.tensor([0, 1,3,1])" ] }, { "cell_type": "code", "execution_count": 181, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "ebd5x51i4CZz", "outputId": "4f3b738c-4c2b-49e9-d09a-44cf83f62bf1" }, "outputs": [ { "data": { "text/plain": [ "tensor([0.0000, 1.1000, 3.3000, 1.1000])" ] }, "execution_count": 181, "metadata": {}, "output_type": "execute_result" } ], "source": [ "a[b]" ] }, { "cell_type": "code", "execution_count": 185, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([False, False, False, True, True, True, False])" ] }, "execution_count": 185, "metadata": {}, "output_type": "execute_result" } ], "source": [ "a > 3" ] }, { "cell_type": "markdown", "metadata": { "id": "jLYmginy4CZz" }, "source": [ "We can find all elements that are greater than 3 and assign them to zero:" ] }, { "cell_type": "code", "execution_count": 186, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Zq-QWUW-4CZz", "outputId": "f68c759a-96c4-401d-f778-8e140fad2a9c" }, "outputs": [ { "data": { "text/plain": [ "tensor([0.0000, 1.1000, 2.2000, 0.0000, 0.0000, 0.0000, 2.2000])" ] }, "execution_count": 186, "metadata": {}, "output_type": "execute_result" } ], "source": [ "a[a>3] = 0\n", "a" ] }, { "cell_type": "markdown", "metadata": { "id": "pfquCanR4CZ0" }, "source": [ "Here t>3 creates a **mask** tensor:" ] }, { "cell_type": "code", "execution_count": 187, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "lEGZzqPE4CZ0", "outputId": "b984e27b-5c57-48df-8bc7-00236f7f4547" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([False, False, False, True, True, True, False])\n" ] } ], "source": [ "a = torch.tensor([0.0, 1.1, 2.2, 3.3, 4.4, 5.5, 2.2])\n", "print(a>3)" ] }, { "cell_type": "markdown", "metadata": { "id": "8JEfa4CZ4CZ1" }, "source": [ "Grab all rows that have a sum greater than 5:" ] }, { "cell_type": "code", "execution_count": 191, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "MPaIchyY4CZ1", "outputId": "967061f3-02a1-4bd0-e9f1-6b6d12070116", "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', requires_grad=True)\n" ] }, { "data": { "text/plain": [ "tensor([[2.],\n", " [4.],\n", " [6.]], device='cuda:0', grad_fn=)" ] }, "execution_count": 191, "metadata": {}, "output_type": "execute_result" } ], "source": [ "print(my_tensor)\n", "my_tensor[:, my_tensor.sum(axis=0) > 10]" ] }, { "cell_type": "code", "execution_count": 195, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', grad_fn=)" ] }, "execution_count": 195, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_tensor[:]" ] }, { "cell_type": "code", "execution_count": 192, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.]], device='cuda:0', requires_grad=True)\n" ] }, { "data": { "text/plain": [ "tensor([[3., 4.],\n", " [5., 6.]], device='cuda:0', grad_fn=)" ] }, "execution_count": 192, "metadata": {}, "output_type": "execute_result" } ], "source": [ "print(my_tensor)\n", "my_tensor[my_tensor.sum(axis=1) > 5]" ] }, { "cell_type": "markdown", "metadata": { "id": "fjEgPVlE4CZ2" }, "source": [ "### 1.6 Broadcasting" ] }, { "cell_type": "markdown", "metadata": { "id": "UZaj_2a84CZ3" }, "source": [ "Sometimes, you can perform operations on tensors of different size via broadcasting. Two tensors can be broadcasted when:\n", "- Each tensor has at least one dimension.\n", "- When iterating over the dimension sizes, starting at the trailing dimension, the dimension sizes must either be:\n", " - equal\n", " - one of them is 1\n", " - one of them does not exist." ] }, { "cell_type": "code", "execution_count": 196, "metadata": { "id": "oU70k1CI4CZ3" }, "outputs": [], "source": [ "t = torch.arange(4)" ] }, { "cell_type": "code", "execution_count": 198, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "90te7sTE4CZ3", "outputId": "9ae7ab39-a6b1-4337-dbc0-8be4a1a07121" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([0, 1, 2, 3])\n", "tensor([100, 101, 102, 103])\n" ] } ], "source": [ "print(t)\n", "print(t + torch.tensor([100]))" ] }, { "cell_type": "markdown", "metadata": { "id": "sKQpiy8m4CZ3" }, "source": [ "Here, we are implicitly broadcasting *tensor([100])*." ] }, { "cell_type": "code", "execution_count": 199, "metadata": { "id": "p_Ot3H3l4CZ3" }, "outputs": [], "source": [ "tensor_2_4 = torch.tensor([[1,2,3,4], [5,6,7,8]], dtype=torch.float32, requires_grad=True)\n", "tensor_4_2 = torch.tensor([[1,2], [3,4], [5,6], [7,8]], dtype=torch.float32, requires_grad=True)\n", "tensor_4 = torch.tensor([.1, .2, .3, .4], dtype=torch.float32, requires_grad=True)" ] }, { "cell_type": "code", "execution_count": 200, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "i9LYNbaw4CZ4", "outputId": "8f5d2ba5-5e32-4530-da87-798c30dfb46d" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor_2_4.shape=torch.Size([2, 4])\n", "tensor([[1., 2., 3., 4.],\n", " [5., 6., 7., 8.]], requires_grad=True)\n", "\n", "tensor_4_2.shape=torch.Size([4, 2])\n", "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.],\n", " [7., 8.]], requires_grad=True)\n", "\n", "tensor_4.shape=torch.Size([4])\n", "tensor([0.1000, 0.2000, 0.3000, 0.4000], requires_grad=True)\n" ] } ], "source": [ "print(f\"{tensor_2_4.shape=}\", tensor_2_4, '', sep='\\n')\n", "print(f\"{tensor_4_2.shape=}\", tensor_4_2, '', sep='\\n')\n", "\n", "print(f\"{tensor_4.shape=}\", tensor_4, sep='\\n')" ] }, { "cell_type": "markdown", "metadata": { "id": "o567g-Ie4CZ5" }, "source": [ "Tensor 2_4 and 4 can be combined because their trailing dimension sizes match:" ] }, { "cell_type": "code", "execution_count": 201, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "_Od4bu_A4CZ5", "outputId": "b66ee8a0-1949-4076-ae44-d61f31cb385f" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1.1000, 2.2000, 3.3000, 4.4000],\n", " [5.1000, 6.2000, 7.3000, 8.4000]], grad_fn=)" ] }, "execution_count": 201, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_2_4 + tensor_4\n", "# size [2, 4] + size [4], okay" ] }, { "cell_type": "markdown", "metadata": { "id": "EorA-lgz4CZ6" }, "source": [ "Trying to add any of the other combinations will result in an error:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "T_-sn-UY4CZ6", "outputId": "d6f6f8d2-df9b-497b-ac8c-125982ce83dd" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "RuntimeError : The size of tensor a (4) must match the size of tensor b (2) at non-singleton dimension 1\n" ] } ], "source": [ "try:\n", " tensor_2_4 + tensor_4_2\n", " # size [2, 4] + size [4, 2], error\n", "except Exception as e:\n", " print(type(e).__name__, \":\", e)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "iDKMXQPy4CZ6", "outputId": "f640d71b-5dbd-4260-a285-b9aaf84b3147" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "RuntimeError : The size of tensor a (2) must match the size of tensor b (4) at non-singleton dimension 1\n" ] } ], "source": [ "try:\n", " tensor_4_2 + tensor_4\n", " # size [4, 2] + size [4], error\n", "except Exception as e:\n", " print(type(e).__name__, \":\", e)" ] }, { "cell_type": "markdown", "metadata": { "id": "h7AwEPqbz0zs" }, "source": [ "You can introduce new dimensions to the existing tensor to make broadcasting possible" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "q8uAwt59zZqp", "outputId": "ffa24f13-a245-4044-a4fa-86b9efd1e1f7" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor_4.shape=torch.Size([4])\n", "tensor_4.unsqueeze(0).shape=torch.Size([1, 4])\n", "tensor_4.unsqueeze(1).shape=torch.Size([4, 1])\n" ] } ], "source": [ "print(f\"{tensor_4.shape=}\")\n", "print(f\"{tensor_4.unsqueeze(0).shape=}\")\n", "print(f\"{tensor_4.unsqueeze(1).shape=}\")" ] }, { "cell_type": "code", "execution_count": 202, "metadata": {}, "outputs": [], "source": [ "tensor_test1 = torch.arange(12).view(2, 2, 3)\n", "tensor_test2 = torch.arange(6).view(2,3)" ] }, { "cell_type": "code", "execution_count": 203, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[[ 0, 1, 2],\n", " [ 3, 4, 5]],\n", "\n", " [[ 6, 7, 8],\n", " [ 9, 10, 11]]])\n", "tensor([[0, 1, 2],\n", " [3, 4, 5]])\n" ] } ], "source": [ "print(tensor_test1, tensor_test2, sep=\"\\n\")" ] }, { "cell_type": "code", "execution_count": 204, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([[[ 0, 2, 4],\n", " [ 3, 5, 7]],\n", "\n", " [[ 9, 11, 13],\n", " [12, 14, 16]]])" ] }, "execution_count": 204, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_test1 + tensor_test2[:, None, :]" ] }, { "cell_type": "code", "execution_count": 205, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([[[ 0, 2, 4],\n", " [ 6, 8, 10]],\n", "\n", " [[ 6, 8, 10],\n", " [12, 14, 16]]])" ] }, "execution_count": 205, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_test1 + tensor_test2[None, :, :]" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "MZOf2u7Tzmqa", "outputId": "b60a8e2f-e261-4da3-ffeb-62434865e35c" }, "outputs": [ { "data": { "text/plain": [ "torch.Size([4, 2])" ] }, "execution_count": 241, "metadata": {}, "output_type": "execute_result" } ], "source": [ "result = tensor_4_2 + tensor_4.unsqueeze(1)\n", "# size [4, 2] + size [4, 1], okay\n", "result.shape" ] }, { "cell_type": "code", "execution_count": 215, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([[0, 1, 2],\n", " [3, 4, 5]])" ] }, "execution_count": 215, "metadata": {}, "output_type": "execute_result" } ], "source": [ "t = torch.arange(6).view(2,3)\n", "t" ] }, { "cell_type": "code", "execution_count": 216, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "tensor([[0, 1],\n", " [2, 3],\n", " [4, 5]])" ] }, "execution_count": 216, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.arange(6).view(-1, 2)" ] }, { "cell_type": "code", "execution_count": 210, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "torch.Size([2, 3, 1])" ] }, "execution_count": 210, "metadata": {}, "output_type": "execute_result" } ], "source": [ "t.unsqueeze(2).size()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "X8zP5t59z7F_", "outputId": "20396118-a1ba-4f3b-a52c-3af9fbc47954" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor_4_2.shape=torch.Size([4, 2])\n", "tensor_4_2.unsqueeze(0).shape=torch.Size([1, 4, 2])\n", "tensor_4_2.unsqueeze(1).shape=torch.Size([4, 1, 2])\n", "tensor_4_2.unsqueeze(2).shape=torch.Size([4, 2, 1])\n", "tensor_4_2.unsqueeze(-3).shape=torch.Size([1, 4, 2])\n", "tensor_4_2.unsqueeze(-2).shape=torch.Size([4, 1, 2])\n", "tensor_4_2.unsqueeze(-1).shape=torch.Size([4, 2, 1])\n" ] } ], "source": [ "print(f\"{tensor_4_2.shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(0).shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(1).shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(2).shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(-3).shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(-2).shape=}\")\n", "print(f\"{tensor_4_2.unsqueeze(-1).shape=}\")" ] }, { "cell_type": "markdown", "metadata": { "id": "xyhKjJEs4CZ6" }, "source": [ "### 1.7 Tensor Manipulation" ] }, { "cell_type": "markdown", "metadata": { "id": "fpW5tpf24CZ8" }, "source": [ "**.view** allows you change the shape of the tensor w/o changing the underlying data:" ] }, { "cell_type": "code", "execution_count": 211, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Lif28MYd4CZ8", "outputId": "300ea49b-2066-43fd-8717-4f78b60b4756" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([2, 4])\n", "tensor([[1., 2., 3., 4.],\n", " [5., 6., 7., 8.]], requires_grad=True)\n" ] } ], "source": [ "print(tensor_2_4.shape, tensor_2_4, sep='\\n')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "9Q9QWsWO4CZ8", "outputId": "4cbc4098-120b-46ce-f65e-6edcd9d006e2" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([4, 2])\n", "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.],\n", " [7., 8.]], grad_fn=)\n" ] } ], "source": [ "print(tensor_2_4.view((4,2)).shape, tensor_2_4.view((4,2)), sep='\\n')" ] }, { "cell_type": "markdown", "metadata": { "id": "DKP4XaPr4CZ8" }, "source": [ "We can add the two tensors now that they are the same shape:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "gULkDCym4CZ8", "outputId": "58544fd8-5659-415f-d7df-7f101f71b95e" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([4, 2])\ttorch.Size([4, 2])\n", "tensor([[ 2., 4.],\n", " [ 6., 8.],\n", " [10., 12.],\n", " [14., 16.]], grad_fn=)\n" ] } ], "source": [ "print(tensor_2_4.view((4,2)).shape, tensor_4_2.shape, sep='\\t')\n", "print(tensor_2_4.view((4,2)) + tensor_4_2)" ] }, { "cell_type": "markdown", "metadata": { "id": "NZ2ig4T94CZ9" }, "source": [ "You can also use -1 to let PyTorch figure out the dimension for you:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "fkMyMVPa4CZ9", "outputId": "eadf653e-3720-4e29-bfb6-d283586c9b1a" }, "outputs": [ { "data": { "text/plain": [ "tensor([[1., 2.],\n", " [3., 4.],\n", " [5., 6.],\n", " [7., 8.]], grad_fn=)" ] }, "execution_count": 246, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_2_4.view((-1,2)) # total 8 elements, so -1, 2 is equivalent to 4, 2" ] }, { "cell_type": "code", "execution_count": 217, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "g_RqgKJH4CZ9", "outputId": "6b666a88-ef0c-496e-f881-2b46c52e6ec8" }, "outputs": [ { "data": { "text/plain": [ "tensor([1., 2., 3., 4., 5., 6., 7., 8.], grad_fn=)" ] }, "execution_count": 217, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tensor_2_4.view(-1)" ] }, { "cell_type": "code", "execution_count": 218, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "CkvAs8wx2bNq", "outputId": "12da13ef-8f11-416e-93e9-d5ee0dbed6d1" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([1., 2., 3., 4., 5., 6., 7., 8.], grad_fn=)\n", "tensor([1., 2., 3., 4., 5., 6., 7., 8.], grad_fn=)\n" ] } ], "source": [ "print(torch.flatten(tensor_2_4))\n", "print(tensor_2_4.flatten())" ] }, { "cell_type": "markdown", "metadata": { "id": "Izqc7sXd4CZ9" }, "source": [ "**torch.cat** concatenates two tensors along the specified dimension:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "n0DcutpQ4CZ9" }, "outputs": [], "source": [ "tensor_4_2_new = tensor_2_4.view((4,2)) * 0.1" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "L14cOhNl4CZ-", "outputId": "2f5b69ae-b427-40de-e4d8-b13db8d0b38f" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([4, 2])\ttorch.Size([4, 2])\n" ] } ], "source": [ "print(tensor_4_2_new.shape, tensor_4_2.shape, sep='\\t')" ] }, { "cell_type": "code", "execution_count": 219, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "YuZCGg_S4CZ-", "outputId": "5ea11f97-58b6-4a1f-dc4d-30352b898640" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[0.1000, 0.2000],\n", " [0.3000, 0.4000],\n", " [0.5000, 0.6000],\n", " [0.7000, 0.8000],\n", " [1.0000, 2.0000],\n", " [3.0000, 4.0000],\n", " [5.0000, 6.0000],\n", " [7.0000, 8.0000]], grad_fn=)\n" ] } ], "source": [ "# 4x2, 4x2 -> 8x2\n", "print(torch.cat([tensor_4_2_new, tensor_4_2], dim=0))" ] }, { "cell_type": "code", "execution_count": 220, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "HeRuukb04CZ-", "outputId": "8316e9ec-53c3-4be0-a3aa-83b9d9118dfd" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([[0.1000, 0.2000, 1.0000, 2.0000],\n", " [0.3000, 0.4000, 3.0000, 4.0000],\n", " [0.5000, 0.6000, 5.0000, 6.0000],\n", " [0.7000, 0.8000, 7.0000, 8.0000]], grad_fn=)\n" ] } ], "source": [ "# 4x2, 4x2 -> 4x4\n", "print(torch.cat([tensor_4_2_new, tensor_4_2], dim=1))" ] }, { "cell_type": "markdown", "metadata": { "id": "taE4ILX14CZ_" }, "source": [ "Note that the tensor dimensions must match along the non-specified dimension. Otherwise, it will throw an error like this:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "6hkYhAfg4CZ_", "outputId": "3a233e47-c403-4cae-9d8e-8e1b22970a66" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "RuntimeError : Sizes of tensors must match except in dimension 1. Expected size 2 but got size 4 for tensor number 1 in the list.\n" ] } ], "source": [ "try:\n", " torch.cat([tensor_2_4, tensor_4_2], dim=1)\n", " # 2x4, 4x2 -> error\n", "except Exception as e:\n", " print(type(e).__name__, \":\", e)" ] }, { "cell_type": "markdown", "metadata": { "id": "Xmo2iSzMkyN1" }, "source": [ "Similarly, **torch.stack** stack a list of tensors (with the same sizes) along a new dimension." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "HYm4GugClmS0", "outputId": "32ad5292-0a74-4478-c768-1c56c2ddb588" }, "outputs": [ { "data": { "text/plain": [ "tensor([[0., 0., 0.],\n", " [1., 1., 1.],\n", " [2., 2., 2.]])" ] }, "execution_count": 257, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.stack([torch.zeros(3), torch.ones(3), torch.ones(3)+1])" ] }, { "cell_type": "code", "execution_count": 225, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "torch.Size([5, 3])" ] }, "execution_count": 225, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.cat([torch.zeros((2,3)), torch.ones((2,3)), torch.ones((1,3))+1], dim=0).size()" ] }, { "cell_type": "code", "execution_count": 223, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "torch.Size([3, 2, 3])" ] }, "execution_count": 223, "metadata": {}, "output_type": "execute_result" } ], "source": [ "torch.stack([torch.zeros((2,3)), torch.ones((2,3)), torch.ones((2,3))+1]).size()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "O3aXqCQq4kSG", "outputId": "cd90331a-cbb5-43f5-cace-75144ac77c46" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor([0., 0., 0., 1., 1., 1., 2., 2., 2.])\n", "tensor([[0., 0., 0.],\n", " [1., 1., 1.],\n", " [2., 2., 2.]])\n" ] } ], "source": [ "print(torch.cat([torch.zeros(3), torch.ones(3), torch.ones(3)+1]))\n", "print(torch.cat([torch.zeros(3).unsqueeze(0), torch.ones(3).unsqueeze(0), torch.ones(3).unsqueeze(0)+1]))" ] }, { "cell_type": "markdown", "metadata": { "id": "TOcd_ufP4CaC" }, "source": [ "## 2. Neural Network Modules" ] }, { "cell_type": "markdown", "metadata": { "id": "x7D1asX_4CaC" }, "source": [ "This section will walk you through some neural network modules relevant to the assignment and NLP in general." ] }, { "cell_type": "markdown", "metadata": { "id": "YW7B_H0y4CaC" }, "source": [ "### 2.1 Embedding" ] }, { "cell_type": "markdown", "metadata": { "id": "PqHOqTcs4CaC" }, "source": [ "**torch.nn.Embedding** is essentially a lookup table for your word embeddings. It keeps track of a V\\*E matrix where V is your vocabulary size and E is the embedding size." ] }, { "cell_type": "markdown", "metadata": { "id": "rEaTA2BMHBhp" }, "source": [ "![pic1.jpeg](data:image/jpeg;base64,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)" ] }, { "cell_type": "markdown", "metadata": { "id": "G26UNmH1HPUi" }, "source": [ "![pic2.jpeg](data:image/jpeg;base64,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)" ] }, { "cell_type": "code", "execution_count": 227, "metadata": { "id": "KJIOmYio4CaC" }, "outputs": [], "source": [ "V = 10\n", "E = 300\n", "embeddings = torch.nn.Embedding(V, E)" ] }, { "cell_type": "code", "execution_count": 228, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "yMDW1C8q4CaD", "outputId": "3798c81f-4403-45a5-c28b-cadca2b5cf86" }, "outputs": [ { "data": { "text/plain": [ "Embedding(10, 300)" ] }, "execution_count": 228, "metadata": {}, "output_type": "execute_result" } ], "source": [ "embeddings" ] }, { "cell_type": "markdown", "metadata": { "id": "T6J1GzfH4CaD" }, "source": [ "Supporse we have pre-processed our sentences into vocabulary indices. We can apply the Embedding object to retrieve the corresponding word embeddings:" ] }, { "cell_type": "code", "execution_count": 234, "metadata": { "id": "9L_yjtAY4CaD" }, "outputs": [], "source": [ "sentences = [[1,2,3], [2,9, 0]]\n", "# \"a\", \"b\", \"c\" ....\n", "# [\"a b c\", \"b i 0\"]" ] }, { "cell_type": "code", "execution_count": 235, "metadata": { "id": "pONPc8Om4CaD" }, "outputs": [], "source": [ "sent_embeds = embeddings(torch.tensor(sentences))" ] }, { "cell_type": "code", "execution_count": 231, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "cQnfZjPx4CaE", "outputId": "a90574fa-6e59-4c5e-9e31-398fa14c5581" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([2, 3, 300])\n", "tensor([[[-0.0316, 0.9131, 0.4405, ..., -0.4678, -1.8660, -0.9187],\n", " [-0.4722, 0.4503, 1.4372, ..., -0.0060, 1.1183, 0.5492],\n", " [-0.4866, 0.6258, -0.3548, ..., 2.3163, 0.9172, 0.4411]],\n", "\n", " [[-0.4722, 0.4503, 1.4372, ..., -0.0060, 1.1183, 0.5492],\n", " [-0.5420, -1.4728, 0.7822, ..., -0.0828, 1.5657, 1.1045],\n", " [-0.5058, -0.5563, -2.3918, ..., 1.6078, 0.3244, -0.3180]]],\n", " grad_fn=)\n" ] } ], "source": [ "print(sent_embeds.size())\n", "print(sent_embeds)" ] }, { "cell_type": "markdown", "metadata": { "id": "Y2l_18ra4CaE" }, "source": [ "Note that 1) the embeddings have been initialized and 2) how the dimensions are being preserved." ] }, { "cell_type": "markdown", "metadata": { "id": "B8WX_1214CaE" }, "source": [ "The resulting tensors will also keep track of gradients now:" ] }, { "cell_type": "code", "execution_count": 236, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "TR12oYfQ4CaE", "outputId": "0708bade-3406-4ca0-bb2a-bee0f858cb84" }, "outputs": [ { "data": { "text/plain": [ "True" ] }, "execution_count": 236, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sent_embeds.requires_grad" ] }, { "cell_type": "markdown", "metadata": { "id": "aA_ro2lJGChW" }, "source": [ "The embedding module has builtin support for padding. You can specify the `padding_idx` parameter." ] }, { "cell_type": "code", "execution_count": 237, "metadata": { "id": "5691EWgYGX0B" }, "outputs": [], "source": [ "embeddings = torch.nn.Embedding(V, E, padding_idx=0)" ] }, { "cell_type": "code", "execution_count": 238, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "j5o7RHTtGbPy", "outputId": "0dd05c9d-e5f1-4aff-b6d0-d7c92228f223" }, "outputs": [ { "data": { "text/plain": [ "tensor([[[ 0.6862, -2.7422, -0.3493, ..., 0.7484, 0.8227, 0.0409],\n", " [ 2.0948, -1.3528, -0.6042, ..., 0.1624, 0.3458, 1.2186],\n", " [ 0.8304, -0.2158, 1.4036, ..., 0.0453, -0.5075, 1.3469]],\n", "\n", " [[ 2.0948, -1.3528, -0.6042, ..., 0.1624, 0.3458, 1.2186],\n", " [ 0.9032, -1.8423, 2.9078, ..., 1.5321, 0.9299, 1.6987],\n", " [ 0.0000, 0.0000, 0.0000, ..., 0.0000, 0.0000, 0.0000]]],\n", " grad_fn=)" ] }, "execution_count": 238, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sent_embeds = embeddings(torch.tensor(sentences))\n", "sent_embeds\n", "# recall sentences = [[1,2,3], [2,9,0]]" ] }, { "cell_type": "markdown", "metadata": { "id": "3Xd6mKtPGnBd" }, "source": [ "Notice how the last row (the last token of the second sentence) is $\\mathbf{0}$." ] }, { "cell_type": "markdown", "metadata": { "id": "4hdpVGuv4CaF" }, "source": [ "### 2.2 Linear Layer" ] }, { "cell_type": "markdown", "metadata": { "id": "pgRzNKeo4CaF" }, "source": [ "**torch.nn.Linear** specifies a fully connected layer. Let's project an embedding to 50 dimensions from 300:" ] }, { "cell_type": "markdown", "metadata": { "id": "YM1GouEXHYKC" }, "source": [ "$$y = x W^\\top + b$$" ] }, { "cell_type": "code", "execution_count": 239, "metadata": { "id": "U5DYyJiW4CaG" }, "outputs": [], "source": [ "linear_layer = torch.nn.Linear(300, 50)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "KsncQgpR4CaG", "outputId": "8646cda1-44f6-400f-f6be-22071aaac960" }, "outputs": [ { "data": { "text/plain": [ "Linear(in_features=300, out_features=50, bias=True)" ] }, "execution_count": 276, "metadata": {}, "output_type": "execute_result" } ], "source": [ "linear_layer" ] }, { "cell_type": "code", "execution_count": 240, "metadata": { "id": "l385nFAw4CaG" }, "outputs": [], "source": [ "projected_embeddings = linear_layer(sent_embeds)" ] }, { "cell_type": "code", "execution_count": 241, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "XwHd5sxz4CaG", "outputId": "83e43356-2e7e-432b-9a12-798ae7c05f9c", "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([2, 3, 300])\ttorch.Size([2, 3, 50])\n" ] } ], "source": [ "print(sent_embeds.shape, projected_embeddings.shape, sep='\\t')" ] }, { "cell_type": "markdown", "metadata": { "id": "hqupVtvh4CaK" }, "source": [ "### 2.3 Useful Modules" ] }, { "cell_type": "markdown", "metadata": { "id": "mcqarKmO4CaL" }, "source": [ "**torch.nn.functional.relu**" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "tw5dHcx84CaL", "outputId": "a10c665b-c4c8-4cbf-8b6b-c778f0381b04", "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor(-1.2725, grad_fn=)\ttensor(1.6307, grad_fn=)\n", "tensor(0., grad_fn=)\ttensor(1.6307, grad_fn=)\n" ] } ], "source": [ "print(projected_embeddings.min(), projected_embeddings.max(), sep='\\t')\n", "print(torch.nn.functional.relu(projected_embeddings).min(), torch.nn.functional.relu(projected_embeddings).max(), sep='\\t')" ] }, { "cell_type": "markdown", "metadata": { "id": "rMFqjVKk4CaN" }, "source": [ "**torch.nn.Dropout** - *p* specifies the dropout rate:" ] }, { "cell_type": "code", "execution_count": 242, "metadata": { "id": "N-1n6y8d4CaN" }, "outputs": [], "source": [ "dropout_layer = torch.nn.Dropout(p=0.9)" ] }, { "cell_type": "code", "execution_count": 243, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "XLq_x0ZT4CaN", "outputId": "be457f2d-a439-4757-8e69-1201ecdb5539", "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "tensor([[[ -0.0000, -0.0000, -0.0000, 0.0000, 3.0286, -0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, -0.0000, 0.0000, 0.0000, 0.0000,\n", " -0.0000, -0.0000, -0.0000, -0.0000, -0.0000, -0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, 0.5835, 0.0000, 0.0000, -0.0000,\n", " 0.0000, 0.0000, -0.0000, 0.0000, 0.0000, -0.0000, 0.0000,\n", " 0.0000, -0.0000, -0.0000, 0.0000, -0.0000, 0.0000, 0.0000,\n", " -0.0000, -0.0000, -0.0000, -0.0000, -5.1797, 0.0000, 0.0000,\n", " 0.0000],\n", " [ -0.0000, -0.0000, -0.0000, 0.0000, 0.0000, -9.2431, -0.0000,\n", " 0.0000, 0.0000, -0.0000, -0.0000, -0.0000, 0.0000, 0.0000,\n", " 0.0000, 0.0000, 0.0000, -11.1992, 0.0000, -0.0000, -0.0000,\n", " 0.0000, -3.0923, 0.0000, -0.0000, -0.0000, -0.0000, -0.0000,\n", " -0.0000, -0.0000, -0.0000, 8.1991, 0.0000, 0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, -0.0000, 0.0000,\n", " 0.0000, -0.0000, 0.0000, 0.0000, 1.6032, 0.0000, -0.0000,\n", " -0.0000],\n", " [ -0.0000, -0.5899, 0.0000, 0.0000, 0.0000, 0.0000, -0.0000,\n", " 0.0000, -0.0000, -0.0000, -0.0000, -0.7888, 0.0000, -0.0000,\n", " -0.0000, 1.5399, -0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n", " 0.0000, 0.0000, -0.0000, -0.0000, -0.2541, -0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, -0.0000, -2.1715,\n", " -0.0000, -0.0000, -0.0000, -0.0000, 0.0000, 0.0000, -8.4997,\n", " -12.5702, 0.0000, 0.0000, -0.0000, 7.8499, 0.0000, 0.0382,\n", " 0.0000]],\n", "\n", " [[ -0.0000, -0.0000, -0.0000, 0.0000, 0.0000, -0.0000, -0.0000,\n", " 0.0000, 0.0000, -0.0000, -0.0000, -0.0000, 0.0000, 0.0000,\n", " 0.0000, 0.0000, 0.0000, -0.0000, 0.0000, -10.6896, -0.0000,\n", " 0.0000, -0.0000, 0.0000, -3.1529, -0.0000, -0.0000, -0.0000,\n", " -0.0000, -0.0000, -0.0000, 0.0000, 0.0000, 0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, -0.0000, 0.0000,\n", " 0.0000, -0.0000, 0.0000, 2.4548, 0.0000, 0.0000, -0.0000,\n", " -0.0000],\n", " [ -0.0000, 0.0000, 0.0000, 0.0000, -0.0000, 0.0000, -0.0000,\n", " -0.0000, 0.0000, -0.0000, 0.0000, -0.0000, 0.0000, 0.0000,\n", " 0.0000, 1.8638, -0.0000, -0.0000, 0.0000, -0.0000, 0.0000,\n", " -0.0000, 0.0000, 2.9471, 0.0000, 0.0000, -0.0000, 0.0000,\n", " -0.8568, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n", " -0.0000, -0.0000, 0.0000, 0.0000, -0.0000, -0.0000, 0.9663,\n", " 0.0000, -0.0000, 0.0000, -0.0000, 0.7441, 0.0000, -0.0000,\n", " -3.1213],\n", " [ -0.0000, 0.0000, -0.0000, 0.0000, 0.0000, -0.0000, 0.0000,\n", " 0.0000, 0.0000, 0.0000, -0.3698, -0.0000, -0.0000, -0.0000,\n", " -0.0000, -0.0000, -0.0000, -0.0000, 0.3355, -0.0000, -0.0000,\n", " 0.0000, 0.0000, 0.0000, -0.0000, -0.0000, 0.0000, -0.4675,\n", " 0.0000, 0.0000, 0.0000, -0.0000, 0.0000, 0.0000, 0.0000,\n", " 0.0000, 0.0000, 0.0000, -0.0000, -0.0000, -0.0000, 0.3695,\n", " -0.0000, -0.1367, -0.0000, -0.0000, 0.0000, 0.0000, 0.0000,\n", " -0.0000]]], grad_fn=)" ] }, "execution_count": 243, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dropout_layer(projected_embeddings)" ] }, { "cell_type": "markdown", "metadata": { "id": "l8EvZGU24CaO" }, "source": [ "Notice how some values are zeroed out.\n", "\n", "Call model.eval() to disable dropout, where model is your model class that inherits nn.Module." ] }, { "cell_type": "markdown", "metadata": { "id": "f455zE024CaQ" }, "source": [ "### 2.4 Training Loop" ] }, { "cell_type": "markdown", "metadata": { "id": "1vJyI_xvN60S" }, "source": [ "Now, let's consider this simple MNIST model." ] }, { "cell_type": "code", "execution_count": 244, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "jg6bHPiBNpXp", "outputId": "d6225ce7-0558-4add-8961-f4a9b1b62f0a" }, "outputs": [], "source": [ "import torch\n", "from torch.utils.data import DataLoader\n", "\n", "import torchvision\n", "import matplotlib.pyplot as plt\n", "\n", "# 0.1 Prepare your data\n", "train_data = torchvision.datasets.MNIST(\n", " root = 'data',\n", " train = True,\n", " transform = torchvision.transforms.ToTensor(),\n", " download=True,\n", ")\n", "test_data = torchvision.datasets.MNIST(\n", " root = 'data',\n", " train = False,\n", " transform = torchvision.transforms.ToTensor()\n", ")\n", "\n", "def show_image(dataset, index):\n", " plt.imshow(dataset[index][0].squeeze(), cmap='gray')\n", " plt.show()\n", "\n", "\n", "train_dataloader = DataLoader(train_data, batch_size=100, shuffle=True)\n", "test_dataloader = DataLoader(test_data, batch_size=100, shuffle=True)\n", "\n", "\n", "# 0.2 Specify your neural network\n", "\n", "class SimpleModel(torch.nn.Module):\n", " def __init__(self, input_size, hidden_size, output_size):\n", " super(SimpleModel, self).__init__()\n", "\n", " self.layer_1 = torch.nn.Linear(input_size, hidden_size)\n", " self.layer_2 = torch.nn.Linear(hidden_size, hidden_size)\n", " self.layer_3 = torch.nn.Linear(hidden_size, output_size)\n", "\n", " self.relu = torch.nn.ReLU()\n", " # self.softmax = torch.nn.LogSoftmax(dim=1)\n", "\n", " def forward(self, batch):\n", "\n", " batch = batch.view(batch.shape[0], -1)\n", "\n", " hidden = self.layer_1(batch)\n", " hidden = self.relu(hidden)\n", " hidden = self.layer_2(hidden)\n", " hidden = self.relu(hidden)\n", " output = self.layer_3(hidden)\n", "\n", " return output" ] }, { "cell_type": "markdown", "metadata": { "id": "vQRXZh8Q4CaQ" }, "source": [ "Below is a quick example of how a training iteration works:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "hHq83NU04CaQ" }, "outputs": [], "source": [ "model = SimpleModel(784, 100, 10)\n", "\n", "# Set up an optimizer and a loss function\n", "optimizer = torch.optim.Adam(model.parameters())\n", "loss_fn = torch.nn.CrossEntropyLoss()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "-SyyynJu4CaR", "outputId": "a9faf63a-fb09-466f-ca14-0f8656460c2d" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "iteration 0: 2.3164162635803223\n", "iteration 1: 2.290942668914795\n", "iteration 2: 2.286985397338867\n", "iteration 3: 2.2663285732269287\n", "iteration 4: 2.247443199157715\n", "iteration 5: 2.2340404987335205\n", "iteration 6: 2.2193119525909424\n", "iteration 7: 2.2141265869140625\n", "iteration 8: 2.1844401359558105\n", "iteration 9: 2.142188310623169\n", "iteration 10: 2.1458661556243896\n", "iteration 11: 2.110956907272339\n", "iteration 12: 2.039005994796753\n", "iteration 13: 2.0035767555236816\n", "iteration 14: 1.971337080001831\n", "iteration 15: 1.8975181579589844\n", "iteration 16: 1.8773837089538574\n", "iteration 17: 1.7911789417266846\n", "iteration 18: 1.8048585653305054\n", "iteration 19: 1.7183837890625\n", "iteration 20: 1.6852025985717773\n" ] } ], "source": [ "# For each training iteration\n", "for i, (batch, target) in enumerate(train_dataloader):\n", "\n", " # Reset back-proped gradient\n", " optimizer.zero_grad()\n", "\n", " # Run your model on data\n", " output = model.forward(batch)\n", "\n", " # Compute loss and back-prop\n", " loss = loss_fn(output, target)\n", " loss.backward()\n", "\n", " # Adjust weights\n", " optimizer.step()\n", "\n", " # Call `loss.item()` to singular value tensor returns its value\n", " print(f\"iteration {i}:\", loss.item())\n", "\n", "\n", " if i >= 20:\n", " break" ] }, { "cell_type": "markdown", "metadata": { "id": "391JL8Vae8sd" }, "source": [ "## 3. Einstein is All You Need: einsum and einop\n", "1. Advanced tools for tensor manipulation.\n", "2. Also suitable for computation.\n", "3. Einsum also available in native numpy and torch.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "3b5RgvlDe2gf" }, "outputs": [], "source": [ "import einops" ] }, { "cell_type": "markdown", "metadata": { "id": "h5sihYQk52qy" }, "source": [ "### 3.1 Tensor Shape Manipulation" ] }, { "cell_type": "code", "execution_count": 245, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "4aoVfAf5fv8x", "outputId": "98484bf0-7417-4358-971e-10f0322b2bb6" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 64, 64])\n" ] } ], "source": [ "# einops: reorder\n", "x = torch.randn(10, 64, 64, 3) # (batch, height, width, channels)\n", "x_reordered = einops.rearrange(x, 'b h w c -> b c h w')\n", "print(x_reordered.shape) # Output: torch.Size([10, 3, 64, 64])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "a-CnImh4f3M6", "outputId": "21e17d06-5dc8-48ec-9ceb-14b84f8cb0c1" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 64, 64])\n" ] } ], "source": [ "# torch: reorder\n", "x = torch.randn(10, 64, 64, 3)\n", "x_reordered = x.permute(0, 3, 1, 2) # Reorder dimensions\n", "print(x_reordered.shape) # Output: torch.Size([10, 3, 64, 64])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "XV1H9YiEf5ql", "outputId": "e9e56c2e-43d7-4e1c-8bba-ef6d37e2c926" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 4096])\n" ] } ], "source": [ "# flatten with einops\n", "x = torch.randn(10, 3, 64, 64) # (batch, channels, height, width)\n", "x_flattened = einops.rearrange(x, 'batch channel height width -> batch channel (height width)')\n", "print(x_flattened.shape) # Output: torch.Size([10, 3, 4096])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "hbSCHD1NgCRb", "outputId": "0337e557-b053-4876-d6f5-2eb642110854" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 4096])\n" ] } ], "source": [ "# flatten with torch\n", "x = torch.randn(10, 3, 64, 64)\n", "x_flattened = x.view(10, 3, -1) # Flatten height and width\n", "print(x_flattened.shape) # Output: torch.Size([10, 3, 4096])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "pV8y_cVPgSOc", "outputId": "38d6269d-b611-4b35-9408-9dacdc01283a" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 64, 8])\n" ] } ], "source": [ "# splitting with einops\n", "x = torch.randn(10, 64 * 8) # (batch, sequence_length * heads)\n", "x_split = einops.rearrange(x, 'b (s h) -> b s h', s=64, h=8)\n", "print(x_split.shape) # Output: torch.Size([10, 64, 8])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "KmaFNKctga7y", "outputId": "a85d19b9-ad28-4fb7-8259-ec6b4072c3ac" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 64, 8])\n" ] } ], "source": [ "# splitting with torch\n", "x = torch.randn(10, 64 * 8)\n", "x_split = x.view(10, 64, 8) # Split the dimensions\n", "print(x_split.shape) # Output: torch.Size([10, 64, 8])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "PiMu0FuagfM_", "outputId": "c7119942-cd5e-45ab-e01f-e64fed1bc8af" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([1, 64, 3072])\n" ] } ], "source": [ "# grouping and slicing with einops\n", "# divide a RGB 256 x 256 image to patches of 32x32, flatten the RGB pixel values in each patch\n", "# there are 64 patches in total\n", "x = torch.randn(1, 3, 256, 256) # (batch, channels, height, width)\n", "# Create patches of size 32x32\n", "patches = einops.rearrange(x, 'b c (h p1) (w p2) -> b (h w) (c p1 p2)', p1=32, p2=32)\n", "print(patches.shape) # Output: torch.Size([1, 64, 3072])" ] }, { "cell_type": "markdown", "metadata": { "id": "p5tIsQ7kggeo" }, "source": [ "### 3.2 Einops for Computation" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "VK-VxAsqglUZ", "outputId": "9960f187-93b2-483c-ef11-6d5be0249ccf" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 64, 64, 3])\n", "torch.Size([10, 64, 3])\n" ] } ], "source": [ "# mean reduction with einops\n", "x = torch.randn(10, 64, 64, 3) # (batch, height, width, channels)\n", "x_reduced = einops.reduce(x, 'b h w c -> b h c', 'mean')\n", "print(x.shape)\n", "print(x_reduced.shape) # Output: torch.Size([10, 64, 3])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "CsUxLCCZgnUv", "outputId": "d0aec512-8494-4bde-a449-19b39b5e9b25" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 64, 3])\n" ] } ], "source": [ "# mean reduction with torch\n", "x = torch.randn(10, 64, 64, 3)\n", "x_reduced = x.mean(dim=2) # Reduce over the width dimension\n", "print(x_reduced.shape) # Output: torch.Size([10, 64, 3])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "vFV2yU26iBXE", "outputId": "d43df156-4dcf-4687-a765-d34211e9b8f9" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 5])\n" ] } ], "source": [ "# matrix multiplication with einops\n", "A = torch.randn(10, 3, 400) # (batch, M, K)\n", "B = torch.randn(10, 400, 5) # (batch, K, N)\n", "C = einops.einsum(A, B, 'b m k, b k n -> b m n') # Batch matrix multiplication\n", "print(C.shape) # Output: torch.Size([10, 3, 5])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "SayYJvykjV5o" }, "outputs": [], "source": [ "# matrix multiplication, flattened\n", "# Initialize the result tensor\n", "C_with_for_loop = torch.zeros(10, 3, 5) # (batch, M, N)\n", "\n", "# Perform batch matrix multiplication using nested for loops\n", "for i in range(A.shape[0]): # Iterate over the batch dimension\n", " for m in range(A.shape[1]): # Iterate over rows of A\n", " for n in range(B.shape[2]): # Iterate over columns of B\n", " sum = 0\n", " for k in range(A.shape[2]): # Iterate over the shared dimension\n", " sum += A[i, m, k] * B[i, k, n] # Multiply and accumulate\n", " C_with_for_loop[i, m, n] = sum # Assign the accumulated sum to the result tensor" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "_7vHKD9ZiNAo", "outputId": "168d0441-82f4-479f-b064-aecd789bd26d" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "torch.Size([10, 3, 5])\n" ] } ], "source": [ "# matrix multiplication with torch\n", "C_torch = torch.bmm(A, B) # Batch matrix multiplication\n", "print(C_torch.shape) # Output: torch.Size([10, 3, 5])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "OzPcPOYt7BLC", "outputId": "c18edb22-546e-4c77-c9c0-5494b61fb045" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "True\tTrue\n" ] } ], "source": [ "print(torch.allclose(C, C_with_for_loop), torch.allclose(C, C_torch), sep='\\t')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "kQ7Wzs1w7J5s", "outputId": "e651ed6e-7653-40da-bf17-b0e0de4bfef3" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "tensor(False)\ttensor(True)\n" ] } ], "source": [ "print((C == C_with_for_loop).all(), (C == C_torch).all(), sep='\\t')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "IJkNkwVO7ej6", "outputId": "325d9f77-e620-4f99-b9d1-220bd03be8cf" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "False\n", "True\n" ] } ], "source": [ "print(torch.allclose(torch.tensor([1.0001]), torch.tensor([1.0])))\n", "print(torch.allclose(torch.tensor([1.0001]), torch.tensor([1.0]), atol=0.1))" ] }, { "cell_type": "markdown", "metadata": { "id": "SE3UOaCzlR_-" }, "source": [ "### 3.3 Einsum in NumPy and PyTorch" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "yRWwDf71kmJW", "outputId": "bcb53d3f-e362-4e99-c6f2-01d48c16a7d5" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "einops:\n", " tensor([[20., 23., 26., 29.],\n", " [56., 68., 80., 92.]])\n", "numpy:\n", " [[20. 23. 26. 29.]\n", " [56. 68. 80. 92.]]\n", "torch:\n", " tensor([[20., 23., 26., 29.],\n", " [56., 68., 80., 92.]])\n" ] } ], "source": [ "# prompt: show a few one liner example of the existence of einsum in einops, native numpy and native torch, and show they lead to the same results. Use very small examples such that the result can be printed (the values, not just the shape)\n", "import torch\n", "import numpy as np\n", "\n", "# Example tensors\n", "A = torch.arange(6).reshape(2, 3).float()\n", "B = torch.arange(12).reshape(3, 4).float()\n", "\n", "# einops\n", "C_einops = einops.einsum(A, B, 'i j, j k -> i k')\n", "print(\"einops:\\n\", C_einops)\n", "\n", "# numpy\n", "A_np = A.numpy()\n", "B_np = B.numpy()\n", "C_numpy = np.einsum('i j, j k -> i k', A_np, B_np)\n", "print(\"numpy:\\n\", C_numpy)\n", "\n", "# torch\n", "C_torch = torch.einsum('i j, j k -> i k', A, B)\n", "print(\"torch:\\n\", C_torch)\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "l3witG1wtktR", "outputId": "df0cb052-11cd-4cf4-cea0-9200ae6adba4" }, "outputs": [ { "data": { "text/plain": [ "tensor([[0., 3.],\n", " [1., 4.],\n", " [2., 5.]])" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" } ], "source": [ "einops.rearrange(A, 'first_dimension second_dimension -> second_dimension first_dimension')" ] }, { "cell_type": "markdown", "metadata": { "id": "m7PGADFSgBcU" }, "source": [] } ], "metadata": { "colab": { "collapsed_sections": [ "Xq2d6RbQu8Yh", "tqkYUUo-hKpd", "YW7B_H0y4CaC", "4hdpVGuv4CaF", "hqupVtvh4CaK", "f455zE024CaQ", "391JL8Vae8sd", "h5sihYQk52qy", "p5tIsQ7kggeo" ], "provenance": [] }, "kernelspec": { "display_name": "Python 3", "name": "python3" }, "language_info": { "name": "python" } }, "nbformat": 4, "nbformat_minor": 0 }