Adds new artifacts colab.

colabs/artifact_basics/Artifact_Basics.ipynb

@@ -0,0 +1,280 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "gsp5VlKQVMud"
+      },
+      "source": [
+        "<a href=\"https://colab.research.google.com/github/wandb/examples/blob/master/colabs/pytorch/Simple_PyTorch_Integration.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>\n",
+        "<!--- @wandbcode{pytorch-video} -->"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "TbzoV1XvVMug"
+      },
+      "source": [
+        "<img src=\"http://wandb.me/logo-im-png\" width=\"400\" alt=\"Weights & Biases\" />\n",
+        "\n",
+        "<!--- @wandbcode{pytorch-video} -->\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "MGA0PLfsVMuh"
+      },
+      "source": [
+        "\n",
+        "Use [Weights & Biases](https://wandb.com) for machine learning experiment tracking, dataset versioning, and project collaboration.\n",
+        "\n",
+        "<div><img /></div>\n",
+        "\n",
+        "<img src=\"https://wandb.me/mini-diagram\" width=\"650\" alt=\"Weights & Biases\" />\n",
+        "\n",
+        "<div><img /></div>\n",
+        "\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "11CLrFgFExMe"
+      },
+      "source": [
+        "\n",
+        "Use W&B Artifacts to track and version data as the inputs and outputs of your W&B Runs. For example, a model training run might take in a dataset as input and trained model as output. In addition to logging hyperparameters and metadata to a run, you can use an artifact to log the dataset used to train the model as input and the resulting model checkpoints as outputs."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "dX5cFz_XCd5h"
+      },
+      "source": [
+        "# Set Up"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "q2EeMdcpC7Dl"
+      },
+      "source": [
+        "In order to log data to our web service, you need to log in and import `wandb`. If this is your first time using W&B, you'll need to sign up for a free account at the link that appears."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "nwEF5d7NChdB"
+      },
+      "outputs": [],
+      "source": [
+        "!pip install wandb\n",
+        "import wandb\n",
+        "wandb.login()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "vp9Zhzp5Hvt1"
+      },
+      "source": [
+        "# Create An Artifact"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "3xEsZzOEnPFY"
+      },
+      "source": [
+        "The general workflow for creating an Artifact is:\n",
+        "\n",
+        "\n",
+        "1.   Intialize a run.\n",
+        "2.   Create an Artifact.\n",
+        "3.   Add a dataset, a model, or another Artifact as an input.\n",
+        "4.   Log the artifact in the W&B platform.\n",
+        "\n",
+        "This can by accomplished in two lines of code:"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "Vndk289hgVVK"
+      },
+      "outputs": [],
+      "source": [
+        "run = wandb.init(project=\"artifact-basics\")\n",
+        "run.log_artifact(artifact_or_path=\"/content/sample_data/mnist_test.csv\", name=\"my_first_artifact\", type=\"dataset\")\n",
+        "run.finish()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "tgzxAogXnwCj"
+      },
+      "source": [
+        "First, initalize the run with `wandb.init()`. In this demo, the code adds the run to the `artifact-basic` project, but you can change the name to anyting you'd like.\n",
+        "\n",
+        "Next, log the artifact with `wandb.artifact()`. In this demo, the artifact is a `dataset` using data from `mnist_test.csv`. You can customize your artifact with a `name` and other metadata- see the Artifacts Reference guide for more information.\n",
+        "\n",
+        "\n",
+        "If you change or add any parameters names, be sure to replicate those changes in the following code samples as well."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "bpGidixAICM1"
+      },
+      "source": [
+        "# Use an Artifact"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "jZC69sSMdX4j"
+      },
+      "source": [
+        "When you need to select a particular Artifact, indicate the artifact you want to use with the `use_artifact` method. This returns a run object. In the proceeding code snippet specifies an artifact called `my_first_artifact` with the alias `latest`:\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "7Ix_ciolJICe"
+      },
+      "outputs": [],
+      "source": [
+        "run = wandb.init(project=\"artifact-basics\")\n",
+        "artifact = run.use_artifact(artifact_or_name=\"my_first_artifact:latest\")\n",
+        "run.finish()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "o-rcMsZYh0Ig"
+      },
+      "source": [
+        "# Download an Artifact"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "u60Orwg5sYuI"
+      },
+      "source": []
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "B2ssTI7LrHS8"
+      },
+      "source": [
+        "To retrieve the path of an Artifact for external use, use the `artifact.download()` method. This will retrieve the directory of an Artifact you select."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "18-7Kd9Oh8cl"
+      },
+      "outputs": [],
+      "source": [
+        "run = wandb.init(project=\"artifact-basics\")\n",
+        "artifact = run.use_artifact(artifact_or_name=\"my_first_artifact:latest\")\n",
+        "# downloads the artifact to a directory on the W&B platform.\n",
+        "datadir = artifact.download()\n",
+        "run.finish()\n",
+        "# prints the current artifact directory\n",
+        "print(u'\\u2500' * 10)\n",
+        "print(\"Data directory located at:\" + datadir)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "sxox0H9grrcx"
+      },
+      "source": [
+        "For more information on ways to customize your Artifact download, see the [Download and Usage guide](https://docs.wandb.ai/guides/artifacts/download-and-use-an-artifact)."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "AyANijPlIN5m"
+      },
+      "source": [
+        "# Update an Artifact"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "e4psfYzGeAVS"
+      },
+      "source": [
+        "You can pass new values to update the `description`, `metadata`, and `alias` of an artifact. Call the `save()` method to update the artifact on the W&B servers. You can update an artifact during or outside a W&B Run.\n",
+        "\n",
+        "\n",
+        "This example changes the `description` of the `my_first_artifact` artifact inside a run:"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "16fCvZUZJD29"
+      },
+      "outputs": [],
+      "source": [
+        "run = wandb.init(project=\"artifact-basics\")\n",
+        "artifact = run.use_artifact(artifact_or_name=\"my_first_artifact:latest\")\n",
+        "artifact.description = \"This is an edited description.\"\n",
+        "artifact.save()\n",
+        "run.finish()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "b6F6b6XHVMus"
+      },
+      "source": [
+        "# Next steps\n",
+        "1. [Artifacts reference documentation](https://docs.wandb.ai/ref/python/artifact): Deep dive into artifact parameters and advanced methods.\n",
+        "2. [Traversing Artifacts](https://docs.wandb.ai/guides/artifacts/explore-and-traverse-an-artifact-graph): Tips on how to work with Artifacts on W&B's web platform.\n",
+        "3. [Artifact Automation](https://docs.wandb.ai/guides/artifacts/project-scoped-automations): Automate downstream actions for versioning artifacts.\n",
+        "4. [Artifacts FAQ](https://docs.wandb.ai/guides/artifacts/artifacts-faqs): Frequently asked questions on Artifact security, workflows, and retention policy."
+      ]
+    }
+  ],
+  "metadata": {
+    "accelerator": "GPU",
+    "colab": {
+      "provenance": []
+    },
+    "kernelspec": {
+      "display_name": "Python 3",
+      "name": "python3"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 0
+}