Merge pull request #59 from truefoundry/js-restore

Restored train-model directory files

train-model/Deploy_your_first_job.ipynb

@@ -0,0 +1,385 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "KPsrbcty6JfS"
+   },
+   "source": [
+    "##### This notebook demonstrates a demo of how you can deploy your first job with TrueFoundry.\n",
+    "\n",
+    "---\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Q9eRIwP76rBv"
+   },
+   "source": [
+    "After you complete the notebook, you will have a successful deployed a job to train a model on the iris dataser. Your jobs deployment dashboard will look like this:\n",
+    "\n",
+    "![](https://files.readme.io/2f6871c-Screenshot_2022-11-16_at_11.43.31_PM.png)\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "RbdHA8sf6qm4"
+   },
+   "source": [
+    "## Project structure\n",
+    "\n",
+    "To complete this guide, you are going to create the following **files**:\n",
+    "\n",
+    "- `train.py` : contains our training code\n",
+    "- `requirements.txt` : contains our dependencies\n",
+    "- `deploy.py` contains our deployment code ( you can also use a deployment configuration for deploying using a YAML file)\n",
+    "\n",
+    "Your **final file structure** is going to look like this:\n",
+    "\n",
+    "```\n",
+    ".\n",
+    "├── train.py\n",
+    "├── deploy.py\n",
+    "└── requirements.txt\n",
+    "```\n",
+    "\n",
+    "As you can see, all the following files are created in the same folder/directory\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "1ag-WSEf642R"
+   },
+   "source": [
+    "# Setup\n",
+    "\n",
+    "Let's first setup all the things we need to deploy our service.\n",
+    "\n",
+    "- Signup or Login on TrueFoundry\n",
+    "- Setup Workspace\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Vr3iFsrv67Mc"
+   },
+   "source": [
+    "Let's start with installing `truefoundry`.\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "x5HP_7Qt6pqw",
+    "outputId": "8bcd94b9-537f-4399-e958-d6e29a3a99c0"
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U \"truefoundry>=0.4.1,<0.5.0\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**Login into TrueFoundry**\n",
+    "\n",
+    "In order to login run the cell below. Host can be found from the TrueFoundry UI as shown below like https://app.truefoundry.com\n",
+    "\n",
+    "![image.png](../common/images/host.png)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "0QL_ZCes6nEz"
+   },
+   "outputs": [],
+   "source": [
+    "!tfy login --host \"<Host name of TrueFoundry UI. e.g. https://company.truefoundry.cloud>\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "5eDM6Z6r6q8G"
+   },
+   "source": [
+    "**Select the `Workspace` in which you want to deploy your application.**\n",
+    "\n",
+    "Once you run the cell below you will get a prompt to enter your Workspace FQN. Follow the docs to\n",
+    "\n",
+    "**Create a Workspace**: https://docs.truefoundry.com/docs/key-concepts#creating-a-workspace\n",
+    "\n",
+    "**Get Existing Workspace FQN**: https://docs.truefoundry.com/docs/key-concepts#getting-workspace-fqn"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import click"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "4AU4ufGu6tpT"
+   },
+   "outputs": [],
+   "source": [
+    "WORKSPACE_FQN = click.prompt(\n",
+    "    \"Enter the Workspace FQN\",\n",
+    "    type=str,\n",
+    ")\n",
+    "print(f\"\\nWorkspace FQN set to {WORKSPACE_FQN!r}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "IgTqSS3Q7J46"
+   },
+   "source": [
+    "# Step 1: Implement the training code\n",
+    "\n",
+    "The first step is to create a job that trains a scikit learn model on iris dataset\n",
+    "\n",
+    "We start with a `train.py` containing our training code and `requirements.txt` with our dependencies.\n",
+    "\n",
+    "```\n",
+    ".\n",
+    "├── train.py\n",
+    "└── requirements.txt\n",
+    "```\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "ewNT1fG07tRY"
+   },
+   "source": [
+    "### **`train.py`**\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "LuU9kOzv7Ci7",
+    "outputId": "002753bd-6b23-48ff-aff6-9fa5c0f380d6"
+   },
+   "outputs": [],
+   "source": [
+    "%%writefile train.py\n",
+    "from sklearn.datasets import load_iris\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "from sklearn.linear_model import LogisticRegression\n",
+    "from sklearn.metrics import classification_report\n",
+    "\n",
+    "# load the dataset\n",
+    "X, y = load_iris(as_frame=True, return_X_y=True)\n",
+    "X = X.rename(columns={\n",
+    "        \"sepal length (cm)\": \"sepal_length\",\n",
+    "        \"sepal width (cm)\": \"sepal_width\",\n",
+    "        \"petal length (cm)\": \"petal_length\",\n",
+    "        \"petal width (cm)\": \"petal_width\",\n",
+    "})\n",
+    "\n",
+    "# NOTE:- You can pass these configurations via command line\n",
+    "# arguments, config file, environment variables.\n",
+    "X_train, X_test, y_train, y_test = train_test_split(\n",
+    "    X, y, test_size=0.2, random_state=42, stratify=y\n",
+    ")\n",
+    "\n",
+    "# Initialize the model\n",
+    "clf = LogisticRegression(solver=\"liblinear\")\n",
+    "# Fit the model\n",
+    "clf.fit(X_train, y_train)\n",
+    "\n",
+    "preds = clf.predict(X_test)\n",
+    "print(classification_report(y_true=y_test, y_pred=preds))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Exa3CI-T7jnR"
+   },
+   "source": [
+    "Click on this [link](https://docs.truefoundry.com/v0.1.1/recipes/training-a-scikit-learn-model) to understand the **`app.py`**:\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "c1zqWidl7yTS"
+   },
+   "source": [
+    "### **`requirements.txt`**\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "vHJts2tR7bfw",
+    "outputId": "7b4c8cbb-2e42-4e6c-e265-d7acf59ffc6a"
+   },
+   "outputs": [],
+   "source": [
+    "%%writefile requirements.txt\n",
+    "pandas==1.5.3\n",
+    "numpy==1.23.2\n",
+    "scikit-learn==1.5.0"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "vtwCrPGJ750K"
+   },
+   "source": [
+    "# Step 2: Deploying as a Job\n",
+    "\n",
+    "You can deploy services on TrueFoundry programmatically via our **Python SDK**.\n",
+    "\n",
+    "Create the `deploy.py`, after which our file structure will look like this:\n",
+    "\n",
+    "**File Structure**\n",
+    "\n",
+    "```Text\n",
+    ".\n",
+    "├── train.py\n",
+    "├── deploy.py\n",
+    "└── requirements.txt\n",
+    "```\n",
+    "\n",
+    "### **`deploy.py`**\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "jc-yRSkE7xdD",
+    "outputId": "ffa6fcc8-f43d-4238-8138-ed5131b2322c"
+   },
+   "outputs": [],
+   "source": [
+    "%%writefile deploy.py\n",
+    "import argparse\n",
+    "import logging\n",
+    "\n",
+    "from truefoundry.deploy import Build, Job, PythonBuild, LocalSource\n",
+    "\n",
+    "logging.basicConfig(\n",
+    "    level=logging.INFO, format=\"%(asctime)s [%(name)s] %(levelname)-8s %(message)s\"\n",
+    ")\n",
+    "\n",
+    "parser = argparse.ArgumentParser()\n",
+    "parser.add_argument(\"--workspace_fqn\", required=True, type=str)\n",
+    "args = parser.parse_args()\n",
+    "\n",
+    "job = Job(\n",
+    "    name=\"iris-train-job\",\n",
+    "    image=Build(\n",
+    "        build_source=LocalSource(local_build=False),\n",
+    "        build_spec=PythonBuild(\n",
+    "            python_version=\"3.11\",\n",
+    "            command=\"python train.py\",\n",
+    "            requirements_path=\"requirements.txt\",\n",
+    "        )\n",
+    "    )\n",
+    ")\n",
+    "job.deploy(workspace_fqn=args.workspace_fqn, wait=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "Ac89hU5U8tUy"
+   },
+   "source": [
+    "Click on this [link](https://docs.truefoundry.com/recipes/deploy-training-code-as-a-job) to understand the **`deploy.py`**:\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "6bjVtz5v829H"
+   },
+   "source": [
+    "Now to deploy our Job run the command below\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "OBNeK70h85L1",
+    "outputId": "58c98c1d-9622-46b3-e977-8eb9e195d22b"
+   },
+   "outputs": [],
+   "source": [
+    "!python deploy.py --workspace_fqn $WORKSPACE_FQN"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "jupyter-base",
+   "language": "python",
+   "name": "jupyter-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}