{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/almutareb/InnovationPathfinderAI/blob/agent_notebook/mixtral_ollama_public_api.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hdHmLCO48n-H"
      },
      "outputs": [],
      "source": [
        "# https://github.com/jmorganca/ollama/blob/main/examples/jupyter-notebook/ollama.ipynb"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "o5Q6OmcN0mkS",
        "outputId": "c7a663c4-b1d8-4354-d1f2-fac8f503f4d3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
            "                                 Dload  Upload   Total   Spent    Left  Speed\n",
            "100  8423    0  8423    0     0   8293      0 --:--:--  0:00:01 --:--:--  8298>>> Downloading ollama...\n",
            "100  8423    0  8423    0     0   7756      0 --:--:--  0:00:01 --:--:--  7763\n",
            "############################################################################################# 100.0%\n",
            ">>> Installing ollama to /usr/local/bin...\n",
            ">>> Creating ollama user...\n",
            ">>> Adding current user to ollama group...\n",
            ">>> Creating ollama systemd service...\n",
            "WARNING: Unable to detect NVIDIA GPU. Install lspci or lshw to automatically detect and install NVIDIA CUDA drivers.\n",
            ">>> The Ollama API is now available at 0.0.0.0:11434.\n",
            ">>> Install complete. Run \"ollama\" from the command line.\n",
            "System has not been booted with systemd as init system (PID 1). Can't operate.\n",
            "Failed to connect to bus: Host is down\n"
          ]
        }
      ],
      "source": [
        "!curl https://ollama.ai/install.sh | sh\n",
        "!command -v systemctl >/dev/null && sudo systemctl stop ollama"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TXtmzrel194r"
      },
      "outputs": [],
      "source": [
        "!pip install aiohttp pyngrok -q"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GLPqSWXT3QBg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a7542b67-3287-4ce5-8459-2e86d8ec7dfd"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": []
        }
      ],
      "source": [
        "from pyngrok import ngrok\n",
        "from google.colab import userdata\n",
        "\n",
        "# https://dashboard.ngrok.com/get-started/your-authtoken\n",
        "NGROK_TOKEN = userdata.get('NGROK_TOKEN')\n",
        "ngrok.set_auth_token(\"NGROK_TOKEN\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "import os\n",
        "import asyncio\n",
        "from aiohttp import ClientSession\n",
        "from google.colab import userdata\n",
        "\n",
        "# Set LD_LIBRARY_PATH so the system NVIDIA library becomes preferred\n",
        "# over the built-in library. This is particularly important for\n",
        "# Google Colab which installs older drivers\n",
        "os.environ.update({'LD_LIBRARY_PATH': '/usr/lib64-nvidia'})\n",
        "NGROK_TOKEN = userdata.get('NGROK_TOKEN')\n",
        "ngrok.set_auth_token(\"NGROK_TOKEN\")\n",
        "\n",
        "async def run(cmd):\n",
        "  '''\n",
        "  run is a helper function to run subcommands asynchronously.\n",
        "  '''\n",
        "  print('>>> starting', *cmd)\n",
        "  p = await asyncio.subprocess.create_subprocess_exec(\n",
        "      *cmd,\n",
        "      stdout=asyncio.subprocess.PIPE,\n",
        "      stderr=asyncio.subprocess.PIPE,\n",
        "  )\n",
        "\n",
        "  async def pipe(lines):\n",
        "    async for line in lines:\n",
        "      print(line.strip().decode('utf-8'))\n",
        "\n",
        "  await asyncio.gather(\n",
        "      pipe(p.stdout),\n",
        "      pipe(p.stderr),\n",
        "  )\n",
        "\n",
        "\n",
        "await asyncio.gather(\n",
        "    run(['ollama', 'serve']),\n",
        "    run(['ngrok', 'config', 'add-authtoken', NGROK_TOKEN]),\n",
        "    run(['ngrok', 'http', '--log', 'stderr', '11434']),\n",
        "    # run([\"ollama\",\"run\",\"llama2\"]),\n",
        "    run([\"ollama\",\"run\",\"mistral\"]),\n",
        ")"
      ],
      "metadata": {
        "id": "u1PG3y83Z-u5"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "gpuType": "T4",
      "authorship_tag": "ABX9TyMLapRySjkHA++gSoAkRSHn",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}