SLM_Lab_Colab.ipynb

slm_lab_colab.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "SLM_Lab_Colab.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/kengz/6fd52a902129fb6d4509c721d71bda48/slm_lab_colab.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nnGf_E2T7d__"
      },
      "source": [
        "# SLM Lab Colab Example\n",
        "\n",
        ">*Although useful for quick exploration, running SLM Lab fully on Colab/Jupyter is not recommended. This notebook is set up for illustrative purposes as requested by a number users.*\n",
        "\n",
        "See https://slm-lab.gitbook.io/slm-lab/setup/installation for the recommended way to setup and run SLM Lab.\n",
        "\n",
        "---\n",
        "*Credit to [@piosif97](https://github.com/piosif97) for the initiative and discussion that led to this notebook.*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gwD6UuUE-a6D"
      },
      "source": [
        "#### 1. Clone the repo and cd into the folder."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "av8zYQQxsGdm"
      },
      "source": [
        "!git clone https://github.com/kengz/SLM-Lab.git"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "147ddSbWtJU8"
      },
      "source": [
        "%cd /content/SLM-Lab"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MfrncRH9-j-1"
      },
      "source": [
        "#### 2. Run the setup script.\n",
        "Note that you'd need to press \"enter\" in the interactive shell in the output cell."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sdE_ZyiGtKQ7"
      },
      "source": [
        "!./bin/setup"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4I6_Ola1-6DL"
      },
      "source": [
        "#### 3. Source the `lab` environment directly and run in headless mode.\n",
        "\n",
        "Since this evaluates as a bash session, you'd need to wait a few minutes for it to finish before Colab outputs the logs here."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NzasgTwN2O91"
      },
      "source": [
        "%%bash\n",
        "# since each shell is a new bash session, this sources the Conda environment directly\n",
        "export PATH=~/miniconda3/envs/lab/bin:$PATH\n",
        "# and we run it in headless mode (Colab has no GUI)\n",
        "# NOTE since each cell evaluates as a session,\n",
        "# the logs will only be printed in the cell output when the command is finished,\n",
        "# i.e. logs don't stream in here, so wait a few minutes to see the output\n",
        "xvfb-run -a python run_lab.py slm_lab/spec/demo.json dqn_cartpole dev"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WIXjhpex4gXd"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}