pytorch_leaning_01_GPU.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0nu6K_S1iO-z"
      },
      "outputs": [],
      "source": [
        "import torch as tc\n",
        "from torch import nn # nn contains all of PyTorch's building blocks for neural networks\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def fun(a,x_):\n",
        "    return a*x_    #$tc.exp(-a*x_**2)\n",
        "\n",
        "# Criando os dados de treino\n",
        "x_train = tc.linspace(-2,2,10).reshape(10,1)\n",
        "y_train = fun(2,x_train)\n",
        "\n",
        "# Criando os dados de teste\n",
        "x_test = tc.linspace(-2,2,40).reshape(40,1)\n",
        "y_test = fun(2,x_test)\n",
        "\n",
        "# Visualizando os dados de treino\n",
        "\n",
        "plt.plot(x_train,y_train,\"k.\" )\n",
        "plt.plot(x_test,y_test,\"r.\",label=\"test\" )\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "H-VOX-EAiYIv",
        "outputId": "00dd6e40-9647-4e5b-e72f-6eab5a71789c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Construindo o modelo\n",
        "\n",
        "class Regressao(nn.Module):\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "        self.linear_layer = nn.Linear(1,1)\n",
        "    def forward(self, x: tc.Tensor) -> tc.Tensor:\n",
        "        return self.linear_layer(x) #"
      ],
      "metadata": {
        "id": "c3giHPOeib9n"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = Regressao()"
      ],
      "metadata": {
        "id": "JJgllLk2jf0Y"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Check model device\n",
        "next(model.parameters()).device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uQWR281piu1_",
        "outputId": "543d6d75-d227-4130-d23e-11ac3a0b305c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cpu')"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Configurando GPU\n",
        "model.to( 'cuda') # the device variable was set above to be \"cuda\" if available or \"cpu\" if not\n",
        "next(model.parameters()).device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XYvedPmui2DP",
        "outputId": "d9707628-7a6d-4558-91a7-25bb94464062"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cuda', index=0)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train = x_train.to( 'cuda')\n",
        "x_test  = x_test.to(  'cuda')\n",
        "y_train = y_train.to( 'cuda')\n",
        "y_test  = y_test.to(  'cuda')\n",
        "\n",
        "loss_fn = nn.L1Loss()\n",
        "optimizer = tc.optim.SGD(params=model.parameters(),lr=0.01)\n",
        "\n",
        "epochs = 1000\n",
        "for epoch in range(epochs):\n",
        "    model.train()\n",
        "\n",
        "    y_preds = model(x_train)\n",
        "    loss = loss_fn(y_preds, y_train)\n",
        "    optimizer.zero_grad()\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    model.eval()\n",
        "\n",
        "    with tc.inference_mode():\n",
        "        y_preds_test = model(x_test)\n",
        "        loss_test = loss_fn(y_preds_test,y_test)\n",
        "        if epoch % 100 == 0:\n",
        "            print(f\"Epoch: {epoch} | MAE Train Loss: {loss} | MAE Test Loss: {loss_test} \")\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Rhet-2jzjPF_",
        "outputId": "8a6b02c4-5574-4635-bc75-ea869a82261b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0 | MAE Train Loss: 2.358814001083374 | MAE Test Loss: 2.183546543121338 \n",
            "Epoch: 100 | MAE Train Loss: 1.1810362339019775 | MAE Test Loss: 1.0853289365768433 \n",
            "Epoch: 200 | MAE Train Loss: 0.08701141923666 | MAE Test Loss: 0.07759565114974976 \n",
            "Epoch: 300 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 400 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 500 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 600 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 700 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 800 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n",
            "Epoch: 900 | MAE Train Loss: 0.007660275790840387 | MAE Test Loss: 0.004369075410068035 \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.eval()\n",
        "with tc.inference_mode():\n",
        "    y_preds = model(x_test)\n",
        "# Visualizando os dados de treino\n",
        "plt.plot(x_test.cpu(),y_preds.cpu(),\".-\" )\n",
        "plt.plot(x_test.cpu(),y_test.cpu(),\"r--\",label=\"test\" )\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "Wzn_efbQkiRf",
        "outputId": "67894bca-2c22-41b7-92a5-c35686440e23"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "N0XmDBfFkeGu"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}