Transcription_with_diarization.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/nahuel89p/whisperx_diarization/blob/main/Transcription_with_diarization.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "3eb38cc7",
      "metadata": {
        "id": "3eb38cc7"
      },
      "source": [
        "# Word by word transcriptions with sepaker diarization 🎙️✨\n",
        "\n",
        "---"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "839c047c",
      "metadata": {
        "id": "839c047c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3390bcfe-e649-4dc6-b848-fcabfacfde06"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting git+https://github.com/m-bain/whisperx.git\n",
            "  Cloning https://github.com/m-bain/whisperx.git to /tmp/pip-req-build-la13z93d\n",
            "  Running command git clone --filter=blob:none --quiet https://github.com/m-bain/whisperx.git /tmp/pip-req-build-la13z93d\n",
            "  Resolved https://github.com/m-bain/whisperx.git to commit b1a98b78c9152ace9f9801593b5fa0c7d5d96b0f\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Collecting pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317 (from whisperx==3.1.1)\n",
            "  Cloning https://github.com/pyannote/pyannote-audio (to revision 11b56a137a578db9335efc00298f6ec1932e6317) to /tmp/pip-install-13443elk/pyannote-audio_68b9046845c343d4b15fea1ce88751b9\n",
            "  Running command git clone --filter=blob:none --quiet https://github.com/pyannote/pyannote-audio /tmp/pip-install-13443elk/pyannote-audio_68b9046845c343d4b15fea1ce88751b9\n",
            "  Running command git rev-parse -q --verify 'sha^11b56a137a578db9335efc00298f6ec1932e6317'\n",
            "  Running command git fetch -q https://github.com/pyannote/pyannote-audio 11b56a137a578db9335efc00298f6ec1932e6317\n",
            "  Running command git checkout -q 11b56a137a578db9335efc00298f6ec1932e6317\n",
            "  Resolved https://github.com/pyannote/pyannote-audio to commit 11b56a137a578db9335efc00298f6ec1932e6317\n",
            "  Running command git submodule update --init --recursive -q\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: torch>=2 in /usr/local/lib/python3.10/dist-packages (from whisperx==3.1.1) (2.0.1+cu118)\n",
            "Requirement already satisfied: torchaudio>=2 in /usr/local/lib/python3.10/dist-packages (from whisperx==3.1.1) (2.0.2+cu118)\n",
            "Collecting faster-whisper (from whisperx==3.1.1)\n",
            "  Downloading faster_whisper-0.8.0-py3-none-any.whl (1.5 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.5/1.5 MB\u001b[0m \u001b[31m11.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting transformers (from whisperx==3.1.1)\n",
            "  Downloading transformers-4.33.1-py3-none-any.whl (7.6 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.6/7.6 MB\u001b[0m \u001b[31m43.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting ffmpeg-python>=0.2 (from whisperx==3.1.1)\n",
            "  Downloading ffmpeg_python-0.2.0-py3-none-any.whl (25 kB)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (from whisperx==3.1.1) (1.5.3)\n",
            "Requirement already satisfied: setuptools>=65 in /usr/local/lib/python3.10/dist-packages (from whisperx==3.1.1) (67.7.2)\n",
            "Requirement already satisfied: nltk in /usr/local/lib/python3.10/dist-packages (from whisperx==3.1.1) (3.8.1)\n",
            "Requirement already satisfied: future in /usr/local/lib/python3.10/dist-packages (from ffmpeg-python>=0.2->whisperx==3.1.1) (0.18.3)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (3.12.2)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (4.5.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (1.12)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (3.1)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (3.1.2)\n",
            "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.10/dist-packages (from torch>=2->whisperx==3.1.1) (2.0.0)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=2->whisperx==3.1.1) (3.27.4.1)\n",
            "Requirement already satisfied: lit in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=2->whisperx==3.1.1) (16.0.6)\n",
            "Collecting av==10.* (from faster-whisper->whisperx==3.1.1)\n",
            "  Downloading av-10.0.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (31.0 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m31.0/31.0 MB\u001b[0m \u001b[31m52.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting ctranslate2<4,>=3.17 (from faster-whisper->whisperx==3.1.1)\n",
            "  Downloading ctranslate2-3.19.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (35.7 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m35.7/35.7 MB\u001b[0m \u001b[31m15.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting huggingface-hub>=0.13 (from faster-whisper->whisperx==3.1.1)\n",
            "  Downloading huggingface_hub-0.17.1-py3-none-any.whl (294 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m294.8/294.8 kB\u001b[0m \u001b[31m31.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting tokenizers==0.13.* (from faster-whisper->whisperx==3.1.1)\n",
            "  Downloading tokenizers-0.13.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (7.8 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.8/7.8 MB\u001b[0m \u001b[31m110.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting onnxruntime<2,>=1.14 (from faster-whisper->whisperx==3.1.1)\n",
            "  Downloading onnxruntime-1.15.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m5.9/5.9 MB\u001b[0m \u001b[31m115.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: click in /usr/local/lib/python3.10/dist-packages (from nltk->whisperx==3.1.1) (8.1.7)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.10/dist-packages (from nltk->whisperx==3.1.1) (1.3.2)\n",
            "Requirement already satisfied: regex>=2021.8.3 in /usr/local/lib/python3.10/dist-packages (from nltk->whisperx==3.1.1) (2023.6.3)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from nltk->whisperx==3.1.1) (4.66.1)\n",
            "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas->whisperx==3.1.1) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas->whisperx==3.1.1) (2023.3.post1)\n",
            "Requirement already satisfied: numpy>=1.21.0 in /usr/local/lib/python3.10/dist-packages (from pandas->whisperx==3.1.1) (1.23.5)\n",
            "Collecting asteroid-filterbanks>=0.4 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading asteroid_filterbanks-0.4.0-py3-none-any.whl (29 kB)\n",
            "Collecting einops>=0.6.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading einops-0.6.1-py3-none-any.whl (42 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m42.2/42.2 kB\u001b[0m \u001b[31m3.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting lightning>=2.0.1 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading lightning-2.0.9-py3-none-any.whl (1.9 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.9/1.9 MB\u001b[0m \u001b[31m61.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting omegaconf<3.0,>=2.1 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading omegaconf-2.3.0-py3-none-any.whl (79 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m79.5/79.5 kB\u001b[0m \u001b[31m9.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pyannote.core>=5.0.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pyannote.core-5.0.0-py3-none-any.whl (58 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m58.5/58.5 kB\u001b[0m \u001b[31m7.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pyannote.database>=5.0.1 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pyannote.database-5.0.1-py3-none-any.whl (48 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m48.1/48.1 kB\u001b[0m \u001b[31m5.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pyannote.metrics>=3.2 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pyannote.metrics-3.2.1-py3-none-any.whl (51 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m51.4/51.4 kB\u001b[0m \u001b[31m5.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pyannote.pipeline>=2.3 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pyannote.pipeline-2.3-py3-none-any.whl (30 kB)\n",
            "Collecting pytorch_metric_learning>=2.1.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pytorch_metric_learning-2.3.0-py3-none-any.whl (115 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m115.3/115.3 kB\u001b[0m \u001b[31m13.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: rich>=12.0.0 in /usr/local/lib/python3.10/dist-packages (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (13.5.2)\n",
            "Collecting semver>=3.0.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading semver-3.0.1-py3-none-any.whl (17 kB)\n",
            "Requirement already satisfied: soundfile>=0.12.1 in /usr/local/lib/python3.10/dist-packages (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.12.1)\n",
            "Collecting speechbrain>=0.5.14 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading speechbrain-0.5.15-py3-none-any.whl (553 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m553.8/553.8 kB\u001b[0m \u001b[31m36.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting tensorboardX>=2.6 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading tensorboardX-2.6.2.2-py2.py3-none-any.whl (101 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m101.7/101.7 kB\u001b[0m \u001b[31m11.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting torch_audiomentations>=0.11.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading torch_audiomentations-0.11.0-py3-none-any.whl (47 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m47.9/47.9 kB\u001b[0m \u001b[31m6.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting torchmetrics>=0.11.0 (from pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading torchmetrics-1.1.2-py3-none-any.whl (764 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m764.8/764.8 kB\u001b[0m \u001b[31m49.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from transformers->whisperx==3.1.1) (23.1)\n",
            "Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from transformers->whisperx==3.1.1) (6.0.1)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers->whisperx==3.1.1) (2.31.0)\n",
            "Collecting safetensors>=0.3.1 (from transformers->whisperx==3.1.1)\n",
            "  Downloading safetensors-0.3.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.3 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.3/1.3 MB\u001b[0m \u001b[31m66.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (2023.6.0)\n",
            "Collecting arrow<3.0,>=1.2.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading arrow-1.2.3-py3-none-any.whl (66 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m66.4/66.4 kB\u001b[0m \u001b[31m8.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting backoff<4.0,>=2.2.1 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading backoff-2.2.1-py3-none-any.whl (15 kB)\n",
            "Requirement already satisfied: beautifulsoup4<6.0,>=4.8.0 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (4.11.2)\n",
            "Collecting croniter<1.5.0,>=1.3.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading croniter-1.4.1-py2.py3-none-any.whl (19 kB)\n",
            "Collecting dateutils<2.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading dateutils-0.6.12-py2.py3-none-any.whl (5.7 kB)\n",
            "Collecting deepdiff<8.0,>=5.7.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading deepdiff-6.5.0-py3-none-any.whl (71 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m71.3/71.3 kB\u001b[0m \u001b[31m8.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting fastapi<2.0,>=0.92.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading fastapi-0.103.1-py3-none-any.whl (66 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m66.2/66.2 kB\u001b[0m \u001b[31m7.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting inquirer<5.0,>=2.10.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading inquirer-3.1.3-py3-none-any.whl (18 kB)\n",
            "Collecting lightning-cloud>=0.5.38 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading lightning_cloud-0.5.38-py3-none-any.whl (659 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m660.0/660.0 kB\u001b[0m \u001b[31m59.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting lightning-utilities<2.0,>=0.7.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading lightning_utilities-0.9.0-py3-none-any.whl (23 kB)\n",
            "Requirement already satisfied: psutil<7.0 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (5.9.5)\n",
            "Requirement already satisfied: pydantic<2.2.0,>=1.7.4 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.10.12)\n",
            "Collecting python-multipart<2.0,>=0.0.5 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading python_multipart-0.0.6-py3-none-any.whl (45 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m45.7/45.7 kB\u001b[0m \u001b[31m3.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting starlette (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading starlette-0.31.1-py3-none-any.whl (69 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m69.9/69.9 kB\u001b[0m \u001b[31m9.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting starsessions<2.0,>=1.2.1 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading starsessions-1.3.0-py3-none-any.whl (10 kB)\n",
            "Requirement already satisfied: traitlets<7.0,>=5.3.0 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (5.7.1)\n",
            "Requirement already satisfied: urllib3<4.0 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.0.4)\n",
            "Collecting uvicorn<2.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading uvicorn-0.23.2-py3-none-any.whl (59 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m59.5/59.5 kB\u001b[0m \u001b[31m7.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: websocket-client<3.0 in /usr/local/lib/python3.10/dist-packages (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.6.2)\n",
            "Collecting websockets<13.0 (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading websockets-11.0.3-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl (129 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m129.9/129.9 kB\u001b[0m \u001b[31m15.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pytorch-lightning (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading pytorch_lightning-2.0.9-py3-none-any.whl (727 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m727.7/727.7 kB\u001b[0m \u001b[31m57.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch>=2->whisperx==3.1.1) (2.1.3)\n",
            "Collecting antlr4-python3-runtime==4.9.* (from omegaconf<3.0,>=2.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading antlr4-python3-runtime-4.9.3.tar.gz (117 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m117.0/117.0 kB\u001b[0m \u001b[31m14.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Collecting coloredlogs (from onnxruntime<2,>=1.14->faster-whisper->whisperx==3.1.1)\n",
            "  Downloading coloredlogs-15.0.1-py2.py3-none-any.whl (46 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m46.0/46.0 kB\u001b[0m \u001b[31m5.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: flatbuffers in /usr/local/lib/python3.10/dist-packages (from onnxruntime<2,>=1.14->faster-whisper->whisperx==3.1.1) (23.5.26)\n",
            "Requirement already satisfied: protobuf in /usr/local/lib/python3.10/dist-packages (from onnxruntime<2,>=1.14->faster-whisper->whisperx==3.1.1) (3.20.3)\n",
            "Requirement already satisfied: sortedcontainers>=2.0.4 in /usr/local/lib/python3.10/dist-packages (from pyannote.core>=5.0.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.4.0)\n",
            "Requirement already satisfied: scipy>=1.1 in /usr/local/lib/python3.10/dist-packages (from pyannote.core>=5.0.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.11.2)\n",
            "Requirement already satisfied: typer[all]>=0.2.1 in /usr/local/lib/python3.10/dist-packages (from pyannote.database>=5.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.9.0)\n",
            "Requirement already satisfied: scikit-learn>=0.17.1 in /usr/local/lib/python3.10/dist-packages (from pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.2.2)\n",
            "Collecting docopt>=0.6.2 (from pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading docopt-0.6.2.tar.gz (25 kB)\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: tabulate>=0.7.7 in /usr/local/lib/python3.10/dist-packages (from pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.9.0)\n",
            "Requirement already satisfied: matplotlib>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.7.1)\n",
            "Collecting optuna>=1.4 (from pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading optuna-3.3.0-py3-none-any.whl (404 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m404.2/404.2 kB\u001b[0m \u001b[31m44.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.1->pandas->whisperx==3.1.1) (1.16.0)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->whisperx==3.1.1) (3.2.0)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->whisperx==3.1.1) (3.4)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->whisperx==3.1.1) (2023.7.22)\n",
            "Requirement already satisfied: markdown-it-py>=2.2.0 in /usr/local/lib/python3.10/dist-packages (from rich>=12.0.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.0.0)\n",
            "Requirement already satisfied: pygments<3.0.0,>=2.13.0 in /usr/local/lib/python3.10/dist-packages (from rich>=12.0.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.16.1)\n",
            "Requirement already satisfied: cffi>=1.0 in /usr/local/lib/python3.10/dist-packages (from soundfile>=0.12.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.15.1)\n",
            "Collecting hyperpyyaml (from speechbrain>=0.5.14->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading HyperPyYAML-1.2.1-py3-none-any.whl (16 kB)\n",
            "Collecting sentencepiece (from speechbrain>=0.5.14->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading sentencepiece-0.1.99-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.3 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.3/1.3 MB\u001b[0m \u001b[31m72.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch>=2->whisperx==3.1.1) (1.3.0)\n",
            "Collecting julius<0.3,>=0.2.3 (from torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading julius-0.2.7.tar.gz (59 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m59.6/59.6 kB\u001b[0m \u001b[31m7.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: librosa>=0.6.0 in /usr/local/lib/python3.10/dist-packages (from torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.10.1)\n",
            "Collecting torch-pitch-shift>=1.2.2 (from torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading torch_pitch_shift-1.2.4-py3-none-any.whl (4.9 kB)\n",
            "Requirement already satisfied: soupsieve>1.2 in /usr/local/lib/python3.10/dist-packages (from beautifulsoup4<6.0,>=4.8.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.5)\n",
            "Requirement already satisfied: pycparser in /usr/local/lib/python3.10/dist-packages (from cffi>=1.0->soundfile>=0.12.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.21)\n",
            "Collecting ordered-set<4.2.0,>=4.0.2 (from deepdiff<8.0,>=5.7.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading ordered_set-4.1.0-py3-none-any.whl (7.6 kB)\n",
            "Requirement already satisfied: anyio<4.0.0,>=3.7.1 in /usr/local/lib/python3.10/dist-packages (from fastapi<2.0,>=0.92.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.7.1)\n",
            "Collecting starlette (from lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading starlette-0.27.0-py3-none-any.whl (66 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m67.0/67.0 kB\u001b[0m \u001b[31m9.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: aiohttp!=4.0.0a0,!=4.0.0a1 in /usr/local/lib/python3.10/dist-packages (from fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (3.8.5)\n",
            "Collecting blessed>=1.19.0 (from inquirer<5.0,>=2.10.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading blessed-1.20.0-py2.py3-none-any.whl (58 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m58.4/58.4 kB\u001b[0m \u001b[31m7.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting python-editor>=1.0.4 (from inquirer<5.0,>=2.10.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading python_editor-1.0.4-py3-none-any.whl (4.9 kB)\n",
            "Collecting readchar>=3.0.6 (from inquirer<5.0,>=2.10.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading readchar-4.0.5-py3-none-any.whl (8.5 kB)\n",
            "Requirement already satisfied: audioread>=2.1.9 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.0.0)\n",
            "Requirement already satisfied: decorator>=4.3.0 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (4.4.2)\n",
            "Requirement already satisfied: numba>=0.51.0 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.56.4)\n",
            "Requirement already satisfied: pooch>=1.0 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.7.0)\n",
            "Requirement already satisfied: soxr>=0.3.2 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.3.6)\n",
            "Requirement already satisfied: lazy-loader>=0.1 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.3)\n",
            "Requirement already satisfied: msgpack>=1.0 in /usr/local/lib/python3.10/dist-packages (from librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.0.5)\n",
            "Requirement already satisfied: pyjwt in /usr/lib/python3/dist-packages (from lightning-cloud>=0.5.38->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.3.0)\n",
            "Requirement already satisfied: mdurl~=0.1 in /usr/local/lib/python3.10/dist-packages (from markdown-it-py>=2.2.0->rich>=12.0.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.1.2)\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.1.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.11.0)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (4.42.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.4.5)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (9.4.0)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=2.0.0->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.1.1)\n",
            "Collecting alembic>=1.5.0 (from optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading alembic-1.12.0-py3-none-any.whl (226 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m226.0/226.0 kB\u001b[0m \u001b[31m26.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting cmaes>=0.10.0 (from optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading cmaes-0.10.0-py3-none-any.whl (29 kB)\n",
            "Collecting colorlog (from optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading colorlog-6.7.0-py2.py3-none-any.whl (11 kB)\n",
            "Requirement already satisfied: sqlalchemy>=1.3.0 in /usr/local/lib/python3.10/dist-packages (from optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.0.20)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.17.1->pyannote.metrics>=3.2->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.2.0)\n",
            "Requirement already satisfied: itsdangerous<3.0.0,>=2.0.1 in /usr/local/lib/python3.10/dist-packages (from starsessions<2.0,>=1.2.1->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.1.2)\n",
            "Collecting primePy>=1.3 (from torch-pitch-shift>=1.2.2->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading primePy-1.3-py3-none-any.whl (4.0 kB)\n",
            "Collecting colorama<0.5.0,>=0.4.3 (from typer[all]>=0.2.1->pyannote.database>=5.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading colorama-0.4.6-py2.py3-none-any.whl (25 kB)\n",
            "Collecting shellingham<2.0.0,>=1.3.0 (from typer[all]>=0.2.1->pyannote.database>=5.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading shellingham-1.5.3-py2.py3-none-any.whl (9.7 kB)\n",
            "Collecting h11>=0.8 (from uvicorn<2.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading h11-0.14.0-py3-none-any.whl (58 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m58.3/58.3 kB\u001b[0m \u001b[31m7.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting humanfriendly>=9.1 (from coloredlogs->onnxruntime<2,>=1.14->faster-whisper->whisperx==3.1.1)\n",
            "  Downloading humanfriendly-10.0-py2.py3-none-any.whl (86 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m86.8/86.8 kB\u001b[0m \u001b[31m10.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting ruamel.yaml<=0.17.28,>=0.17.8 (from hyperpyyaml->speechbrain>=0.5.14->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading ruamel.yaml-0.17.28-py3-none-any.whl (109 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m109.6/109.6 kB\u001b[0m \u001b[31m13.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (23.1.0)\n",
            "Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (6.0.4)\n",
            "Requirement already satisfied: async-timeout<5.0,>=4.0.0a3 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (4.0.3)\n",
            "Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (1.9.2)\n",
            "Requirement already satisfied: frozenlist>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (1.4.0)\n",
            "Requirement already satisfied: aiosignal>=1.1.2 in /usr/local/lib/python3.10/dist-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec->huggingface-hub>=0.13->faster-whisper->whisperx==3.1.1) (1.3.1)\n",
            "Collecting Mako (from alembic>=1.5.0->optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading Mako-1.2.4-py3-none-any.whl (78 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m78.7/78.7 kB\u001b[0m \u001b[31m9.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: sniffio>=1.1 in /usr/local/lib/python3.10/dist-packages (from anyio<4.0.0,>=3.7.1->fastapi<2.0,>=0.92.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.3.0)\n",
            "Requirement already satisfied: exceptiongroup in /usr/local/lib/python3.10/dist-packages (from anyio<4.0.0,>=3.7.1->fastapi<2.0,>=0.92.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (1.1.3)\n",
            "Requirement already satisfied: wcwidth>=0.1.4 in /usr/local/lib/python3.10/dist-packages (from blessed>=1.19.0->inquirer<5.0,>=2.10.0->lightning>=2.0.1->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.2.6)\n",
            "Requirement already satisfied: llvmlite<0.40,>=0.39.0dev0 in /usr/local/lib/python3.10/dist-packages (from numba>=0.51.0->librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (0.39.1)\n",
            "Requirement already satisfied: platformdirs>=2.5.0 in /usr/local/lib/python3.10/dist-packages (from pooch>=1.0->librosa>=0.6.0->torch_audiomentations>=0.11.0->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (3.10.0)\n",
            "Collecting ruamel.yaml.clib>=0.2.7 (from ruamel.yaml<=0.17.28,>=0.17.8->hyperpyyaml->speechbrain>=0.5.14->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1)\n",
            "  Downloading ruamel.yaml.clib-0.2.7-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.manylinux_2_24_x86_64.whl (485 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m485.6/485.6 kB\u001b[0m \u001b[31m36.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: greenlet!=0.4.17 in /usr/local/lib/python3.10/dist-packages (from sqlalchemy>=1.3.0->optuna>=1.4->pyannote.pipeline>=2.3->pyannote.audio@ git+https://github.com/pyannote/pyannote-audio@11b56a137a578db9335efc00298f6ec1932e6317->whisperx==3.1.1) (2.0.2)\n",
            "Building wheels for collected packages: whisperx, pyannote.audio, antlr4-python3-runtime, docopt, julius\n",
            "  Building wheel for whisperx (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for whisperx: filename=whisperx-3.1.1-py3-none-any.whl size=36033 sha256=3d81b7a192c83d2394003a6eac9d8969f00f37c6fb32b63ffd59ed3c1d8fcf40\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-lcc9ovx3/wheels/27/fb/53/682b85073a466f1866910d7257233e53b0cc126ab50e7c5373\n",
            "  Building wheel for pyannote.audio (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pyannote.audio: filename=pyannote.audio-2.1.1-py2.py3-none-any.whl size=400150 sha256=2b43e974c3f407f28fa24a07b8e12ea670bb67fa3830d63012026e4527000c43\n",
            "  Stored in directory: /root/.cache/pip/wheels/1e/f2/4f/ffd6fd3ec96beb935d2f006678342850a3527eca2a9e522df3\n",
            "  Building wheel for antlr4-python3-runtime (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for antlr4-python3-runtime: filename=antlr4_python3_runtime-4.9.3-py3-none-any.whl size=144554 sha256=0e4b4fcc0e92051a6502579165e4b2ea2236a3889fce69cd8e6a290e4fba6e9d\n",
            "  Stored in directory: /root/.cache/pip/wheels/12/93/dd/1f6a127edc45659556564c5730f6d4e300888f4bca2d4c5a88\n",
            "  Building wheel for docopt (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for docopt: filename=docopt-0.6.2-py2.py3-none-any.whl size=13705 sha256=50237e5039761abf77eb9aa8461f0ab3e4ed83e46b5229922b366727b56abb6f\n",
            "  Stored in directory: /root/.cache/pip/wheels/fc/ab/d4/5da2067ac95b36618c629a5f93f809425700506f72c9732fac\n",
            "  Building wheel for julius (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for julius: filename=julius-0.2.7-py3-none-any.whl size=21868 sha256=603151e32648237459ef75786ae2ccd016acb2454e1b3ac0164bb3b795553569\n",
            "  Stored in directory: /root/.cache/pip/wheels/b9/b2/05/f883527ffcb7f2ead5438a2c23439aa0c881eaa9a4c80256f4\n",
            "Successfully built whisperx pyannote.audio antlr4-python3-runtime docopt julius\n",
            "Installing collected packages: tokenizers, sentencepiece, safetensors, python-editor, primePy, docopt, av, antlr4-python3-runtime, websockets, tensorboardX, shellingham, semver, ruamel.yaml.clib, readchar, python-multipart, ordered-set, omegaconf, Mako, lightning-utilities, humanfriendly, h11, ffmpeg-python, einops, ctranslate2, colorlog, colorama, cmaes, blessed, backoff, uvicorn, starlette, ruamel.yaml, pyannote.core, inquirer, huggingface-hub, deepdiff, dateutils, croniter, coloredlogs, arrow, alembic, transformers, starsessions, optuna, onnxruntime, hyperpyyaml, fastapi, pyannote.database, lightning-cloud, faster-whisper, pyannote.pipeline, pyannote.metrics, torchmetrics, torch-pitch-shift, pytorch-lightning, julius, torch_audiomentations, speechbrain, pytorch_metric_learning, lightning, asteroid-filterbanks, pyannote.audio, whisperx\n",
            "Successfully installed Mako-1.2.4 alembic-1.12.0 antlr4-python3-runtime-4.9.3 arrow-1.2.3 asteroid-filterbanks-0.4.0 av-10.0.0 backoff-2.2.1 blessed-1.20.0 cmaes-0.10.0 colorama-0.4.6 coloredlogs-15.0.1 colorlog-6.7.0 croniter-1.4.1 ctranslate2-3.19.0 dateutils-0.6.12 deepdiff-6.5.0 docopt-0.6.2 einops-0.6.1 fastapi-0.103.1 faster-whisper-0.8.0 ffmpeg-python-0.2.0 h11-0.14.0 huggingface-hub-0.17.1 humanfriendly-10.0 hyperpyyaml-1.2.1 inquirer-3.1.3 julius-0.2.7 lightning-2.0.9 lightning-cloud-0.5.38 lightning-utilities-0.9.0 omegaconf-2.3.0 onnxruntime-1.15.1 optuna-3.3.0 ordered-set-4.1.0 primePy-1.3 pyannote.audio-2.1.1 pyannote.core-5.0.0 pyannote.database-5.0.1 pyannote.metrics-3.2.1 pyannote.pipeline-2.3 python-editor-1.0.4 python-multipart-0.0.6 pytorch-lightning-2.0.9 pytorch_metric_learning-2.3.0 readchar-4.0.5 ruamel.yaml-0.17.28 ruamel.yaml.clib-0.2.7 safetensors-0.3.3 semver-3.0.1 sentencepiece-0.1.99 shellingham-1.5.3 speechbrain-0.5.15 starlette-0.27.0 starsessions-1.3.0 tensorboardX-2.6.2.2 tokenizers-0.13.3 torch-pitch-shift-1.2.4 torch_audiomentations-0.11.0 torchmetrics-1.1.2 transformers-4.33.1 uvicorn-0.23.2 websockets-11.0.3 whisperx-3.1.1\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "pydevd_plugins"
                ]
              }
            }
          },
          "metadata": {}
        }
      ],
      "source": [
        "!pip install git+https://github.com/m-bain/whisperx.git --upgrade"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "568bde2d",
      "metadata": {
        "id": "568bde2d"
      },
      "source": [
        "## Upload mp4 files 🎵\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "0fdeeda0",
      "metadata": {
        "id": "0fdeeda0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 355
        },
        "outputId": "26ac63de-c00b-4d29-d54a-1a0ac3d7a697"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-d90358b0-c6a8-4e63-889f-93ae86dcd99b\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-d90358b0-c6a8-4e63-889f-93ae86dcd99b\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script>// Copyright 2017 Google LLC\n",
              "//\n",
              "// Licensed under the Apache License, Version 2.0 (the \"License\");\n",
              "// you may not use this file except in compliance with the License.\n",
              "// You may obtain a copy of the License at\n",
              "//\n",
              "//      http://www.apache.org/licenses/LICENSE-2.0\n",
              "//\n",
              "// Unless required by applicable law or agreed to in writing, software\n",
              "// distributed under the License is distributed on an \"AS IS\" BASIS,\n",
              "// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
              "// See the License for the specific language governing permissions and\n",
              "// limitations under the License.\n",
              "\n",
              "/**\n",
              " * @fileoverview Helpers for google.colab Python module.\n",
              " */\n",
              "(function(scope) {\n",
              "function span(text, styleAttributes = {}) {\n",
              "  const element = document.createElement('span');\n",
              "  element.textContent = text;\n",
              "  for (const key of Object.keys(styleAttributes)) {\n",
              "    element.style[key] = styleAttributes[key];\n",
              "  }\n",
              "  return element;\n",
              "}\n",
              "\n",
              "// Max number of bytes which will be uploaded at a time.\n",
              "const MAX_PAYLOAD_SIZE = 100 * 1024;\n",
              "\n",
              "function _uploadFiles(inputId, outputId) {\n",
              "  const steps = uploadFilesStep(inputId, outputId);\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  // Cache steps on the outputElement to make it available for the next call\n",
              "  // to uploadFilesContinue from Python.\n",
              "  outputElement.steps = steps;\n",
              "\n",
              "  return _uploadFilesContinue(outputId);\n",
              "}\n",
              "\n",
              "// This is roughly an async generator (not supported in the browser yet),\n",
              "// where there are multiple asynchronous steps and the Python side is going\n",
              "// to poll for completion of each step.\n",
              "// This uses a Promise to block the python side on completion of each step,\n",
              "// then passes the result of the previous step as the input to the next step.\n",
              "function _uploadFilesContinue(outputId) {\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  const steps = outputElement.steps;\n",
              "\n",
              "  const next = steps.next(outputElement.lastPromiseValue);\n",
              "  return Promise.resolve(next.value.promise).then((value) => {\n",
              "    // Cache the last promise value to make it available to the next\n",
              "    // step of the generator.\n",
              "    outputElement.lastPromiseValue = value;\n",
              "    return next.value.response;\n",
              "  });\n",
              "}\n",
              "\n",
              "/**\n",
              " * Generator function which is called between each async step of the upload\n",
              " * process.\n",
              " * @param {string} inputId Element ID of the input file picker element.\n",
              " * @param {string} outputId Element ID of the output display.\n",
              " * @return {!Iterable<!Object>} Iterable of next steps.\n",
              " */\n",
              "function* uploadFilesStep(inputId, outputId) {\n",
              "  const inputElement = document.getElementById(inputId);\n",
              "  inputElement.disabled = false;\n",
              "\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  outputElement.innerHTML = '';\n",
              "\n",
              "  const pickedPromise = new Promise((resolve) => {\n",
              "    inputElement.addEventListener('change', (e) => {\n",
              "      resolve(e.target.files);\n",
              "    });\n",
              "  });\n",
              "\n",
              "  const cancel = document.createElement('button');\n",
              "  inputElement.parentElement.appendChild(cancel);\n",
              "  cancel.textContent = 'Cancel upload';\n",
              "  const cancelPromise = new Promise((resolve) => {\n",
              "    cancel.onclick = () => {\n",
              "      resolve(null);\n",
              "    };\n",
              "  });\n",
              "\n",
              "  // Wait for the user to pick the files.\n",
              "  const files = yield {\n",
              "    promise: Promise.race([pickedPromise, cancelPromise]),\n",
              "    response: {\n",
              "      action: 'starting',\n",
              "    }\n",
              "  };\n",
              "\n",
              "  cancel.remove();\n",
              "\n",
              "  // Disable the input element since further picks are not allowed.\n",
              "  inputElement.disabled = true;\n",
              "\n",
              "  if (!files) {\n",
              "    return {\n",
              "      response: {\n",
              "        action: 'complete',\n",
              "      }\n",
              "    };\n",
              "  }\n",
              "\n",
              "  for (const file of files) {\n",
              "    const li = document.createElement('li');\n",
              "    li.append(span(file.name, {fontWeight: 'bold'}));\n",
              "    li.append(span(\n",
              "        `(${file.type || 'n/a'}) - ${file.size} bytes, ` +\n",
              "        `last modified: ${\n",
              "            file.lastModifiedDate ? file.lastModifiedDate.toLocaleDateString() :\n",
              "                                    'n/a'} - `));\n",
              "    const percent = span('0% done');\n",
              "    li.appendChild(percent);\n",
              "\n",
              "    outputElement.appendChild(li);\n",
              "\n",
              "    const fileDataPromise = new Promise((resolve) => {\n",
              "      const reader = new FileReader();\n",
              "      reader.onload = (e) => {\n",
              "        resolve(e.target.result);\n",
              "      };\n",
              "      reader.readAsArrayBuffer(file);\n",
              "    });\n",
              "    // Wait for the data to be ready.\n",
              "    let fileData = yield {\n",
              "      promise: fileDataPromise,\n",
              "      response: {\n",
              "        action: 'continue',\n",
              "      }\n",
              "    };\n",
              "\n",
              "    // Use a chunked sending to avoid message size limits. See b/62115660.\n",
              "    let position = 0;\n",
              "    do {\n",
              "      const length = Math.min(fileData.byteLength - position, MAX_PAYLOAD_SIZE);\n",
              "      const chunk = new Uint8Array(fileData, position, length);\n",
              "      position += length;\n",
              "\n",
              "      const base64 = btoa(String.fromCharCode.apply(null, chunk));\n",
              "      yield {\n",
              "        response: {\n",
              "          action: 'append',\n",
              "          file: file.name,\n",
              "          data: base64,\n",
              "        },\n",
              "      };\n",
              "\n",
              "      let percentDone = fileData.byteLength === 0 ?\n",
              "          100 :\n",
              "          Math.round((position / fileData.byteLength) * 100);\n",
              "      percent.textContent = `${percentDone}% done`;\n",
              "\n",
              "    } while (position < fileData.byteLength);\n",
              "  }\n",
              "\n",
              "  // All done.\n",
              "  yield {\n",
              "    response: {\n",
              "      action: 'complete',\n",
              "    }\n",
              "  };\n",
              "}\n",
              "\n",
              "scope.google = scope.google || {};\n",
              "scope.google.colab = scope.google.colab || {};\n",
              "scope.google.colab._files = {\n",
              "  _uploadFiles,\n",
              "  _uploadFilesContinue,\n",
              "};\n",
              "})(self);\n",
              "</script> "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving QV-69w-51zc.mp4 to QV-69w-51zc.mp4\n",
            "Saving VQHPDtX358Y.mp4 to VQHPDtX358Y.mp4\n",
            "Saving MemAKOLtiGM.mp4 to MemAKOLtiGM.mp4\n",
            "Saving LtUD98tkLPw.mp4 to LtUD98tkLPw.mp4\n",
            "Saving epIFO7OheTI.mp4 to epIFO7OheTI.mp4\n",
            "Saving xDZOQMYRw6w.mp4 to xDZOQMYRw6w.mp4\n",
            "Saving Tr6kX-1wvgM.mp4 to Tr6kX-1wvgM.mp4\n",
            "Saving buS-n3wh6vg.mp4 to buS-n3wh6vg.mp4\n",
            "Saving x1R-VTMknTo.mp4 to x1R-VTMknTo.mp4\n"
          ]
        }
      ],
      "source": [
        "from google.colab import files\n",
        "\n",
        "uploaded = files.upload()\n",
        "audio_file = list(uploaded.keys())[0]"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "57f2303f",
      "metadata": {
        "id": "57f2303f"
      },
      "source": [
        "## Transcription 🗣️\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "86825fe3",
      "metadata": {
        "id": "86825fe3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 254,
          "referenced_widgets": [
            "93ceba654c5245d9ab74eef6e4c617f5",
            "780577efa4634b6fb16c9119cdda58ec",
            "b0dcc88103a8416eab216b362155db83",
            "16bfcd8db8674baca30f989d6f2fb3da",
            "0bf38163d89c4f84bed4a1066231b752",
            "20f35a634fa5492eb3a77ecc2a3dc44f",
            "d00f910737694a0db872991055764be4",
            "b80d79598f0a43bbbcf751d44b977c4f",
            "eca6f18b26bf4a3b874962b3df8a3070",
            "f53b929331f94d70a2f0b727948f2e66",
            "e8bd1a03cdc943e2bb0ac36905054911",
            "9b5c9787f8d346caaff7e8f536704f4b",
            "b93529f8a3bf4eac8d0de82ee8ee8325",
            "322411b766ae4dcca07aa80f3b9cc308",
            "ce1610b9833d43d985d671dd36faa7ea",
            "cb5e7a767c57466e9f5c1a26b56d5853",
            "5cdfaeb9f2dd403b9d1cf844e9695dae",
            "add074b8ec6143d28dc24de836656941",
            "d6559426c08045ab867bf849204124d8",
            "ebda834cc20a4e7381c2c0665b5dc33b",
            "33d0015ff67d4162902af4695b35ee53",
            "104fa4bc207c4ccd82f2cffbabcc916c",
            "0625c2337af64c6fa2e33d58a614e0e8",
            "135589fbf8ec402aad213b5ce38f82df",
            "76688a3bcc4443e583e841f6586544a5",
            "f2fa15344321437bb2de6fcae3785b4f",
            "ed86c9845d4d40708d3dc8ae8c11d52b",
            "a55c3d810b49482d8487668eeaba72ee",
            "1960d9e9f48045f0ba64ee34b8b7927d",
            "d2e2dd6afb8d41b782e149cb2b4defa5",
            "d1df5de8540345feb757b6256b6d0128",
            "938ac00af36b4d86b04e406974b2c00d",
            "6abaf16e6cd341e2900cb28c82e23f65",
            "f55c189b5c9e449696fa9e7a09ccabfb",
            "a1348e8aa1b047608ae14b214dc6d978",
            "5321f147d82843d9a21d74e75153d644",
            "ffa9403b2d984ee7b810a7dee86ce21d",
            "2e3fb4ce3bf4482db12ddc6102f56ad4",
            "86d08e9b45c5430f9bd35bfbad249ad5",
            "812ab7be10bc44b3a948aa14ebab46b2",
            "7b38c06843b94dc696d95d0ac9af6ae8",
            "d25a36bb87ce41ae948b7d3d204a6c95",
            "47161a11268f4b67923e65d44b2f15f0",
            "8f4fd993adfd44359250c5a6eed3885f"
          ]
        },
        "outputId": "c5228030-56f1-4b2b-a3a3-7756061eec96"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Downloading (…)08837e8b/config.json:   0%|          | 0.00/2.80k [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "93ceba654c5245d9ab74eef6e4c617f5"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Downloading (…)37e8b/vocabulary.txt:   0%|          | 0.00/460k [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "9b5c9787f8d346caaff7e8f536704f4b"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Downloading model.bin:   0%|          | 0.00/3.09G [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "0625c2337af64c6fa2e33d58a614e0e8"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Downloading (…)37e8b/tokenizer.json:   0%|          | 0.00/2.20M [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "f55c189b5c9e449696fa9e7a09ccabfb"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "No language specified, language will be first be detected for each audio file (increases inference time).\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████████████████████████████████| 16.9M/16.9M [00:00<00:00, 116MiB/s]\n",
            "INFO:pytorch_lightning.utilities.migration.utils:Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v2.0.9. To apply the upgrade to your files permanently, run `python -m pytorch_lightning.utilities.upgrade_checkpoint --file ../root/.cache/torch/whisperx-vad-segmentation.bin`\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model was trained with pyannote.audio 0.0.1, yours is 2.1.1. Bad things might happen unless you revert pyannote.audio to 0.x.\n",
            "Model was trained with torch 1.10.0+cu102, yours is 2.0.1+cu118. Bad things might happen unless you revert torch to 1.x.\n"
          ]
        }
      ],
      "source": [
        "import whisperx\n",
        "import gc\n",
        "from google.colab import files\n",
        "\n",
        "filenames=list(uploaded.keys())\n",
        "results = []\n",
        "device = \"cuda\"\n",
        "batch_size = 16\n",
        "compute_type = \"float16\"\n",
        "model = whisperx.load_model(\"large-v2\", device, compute_type=compute_type)\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "hf_token = ''\n",
        "for audio_file in filenames:\n",
        "  # Transcript\n",
        "  audio = whisperx.load_audio(audio_file)\n",
        "  result = model.transcribe(audio, batch_size=batch_size)\n",
        "  print(result[\"segments\"])\n",
        "\n",
        "  # Align\n",
        "  model_a, metadata = whisperx.load_align_model(language_code=result[\"language\"], device=device)\n",
        "  result = whisperx.align(result[\"segments\"], model_a, metadata, audio, device, return_char_alignments=False)\n",
        "\n",
        "  diarize_model = whisperx.DiarizationPipeline(use_auth_token=hftoken, device=device)\n",
        "\n",
        "  # add min/max number of speakers if known\n",
        "  diarize_segments = diarize_model(audio) #auto\n",
        "  # diarize_model(audio, min_speakers=min_speakers, max_speakers=max_speakers)\n",
        "\n",
        "  result = whisperx.assign_word_speakers(diarize_segments, result)\n",
        "  import json\n",
        "  file_name= str(audio_file.split('.')[0])+'.json'\n",
        "  # Saving the entire result as a JSON file\n",
        "  with open(file_name, 'w') as file:\n",
        "      json.dump(result, file, indent=4)\n",
        "\n",
        "  # Download the JSON file\n",
        "  files.download(file_name)\n",
        "  results.append(result)\n"
      ],
      "metadata": {
        "id": "mv5vHi822Jkw",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 216
        },
        "outputId": "814ba958-19cf-4e96-bf49-6fb03f86e4d3"
      },
      "id": "mv5vHi822Jkw",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Detected language: en (1.00) in first 30s of audio...\n",
            "[{'text': \" Welcome to Palisades Gold Radio. I'm your host, Tom Bodrovics. Joining me today is Patrick Karim from NorthStarBadShorts.com. How are you today, Pat? I'm doing great. Always fun catching up with you there, especially our conversations even before we start the show, right? I'm telling you, one day, as soon as we hook up, I said, hold on, Patrick, we should record this, show people.\", 'start': 1.477, 'end': 23.397}, {'text': \" all the discussions before, because it's just, I don't know, it's just like a fun conversation there. I feel like, you know, if we make me meet up one day, it's just going to be like, hey, Tom, let's go grab a beer. Let's go have fun. And look, I guess Zoom is the next best thing we have there, but I'll go to your neck of the woods for sure one day there. It sure makes that whatever, seven or what would it be like 5,000 mile commute a little bit easier, but\", 'start': 23.566, 'end': 50.633}, {'text': \" I agree, Pat. There's always the offline conversation, not just with you, but with a lot of guests always seems to be really interesting. But, you know, it's unfortunate that we really have to self-censor ourselves because\", 'start': 50.819, 'end': 67.947}, {'text': \" a lot of the opinions that we end up talking about and might not even be an opinion, but you know, it can be just something that we're thinking about, something that we're questioning ends up getting us flagged, getting us less views. At the end of the day, it's not about, you know, this channel isn't about getting maximizing views. It's about trying to look at all these problems from different\", 'start': 68.858, 'end': 94.475}, {'text': \" different vantage points and trying to, as I say, triangulate our point in the world from all of these different opinions. But unfortunately, that isn't always that palatable.\", 'start': 95.842, 'end': 113.83}, {'text': \" What? There's censorship? What? What are you talking about? This is a free, open platform. What? I thought I could say anything, but you know it exists, censorship, and I'll tell you why. Just the fact that certain words I want to say right now, and I'm holding back, I want to make a joke about something, just the fact I'm holding back, you kind of say it's like my liberty. Why am I doing this? Me, Patrick, if I knew France, it'd be bonkers, the stuff I'd say, you know? But now, yeah, it's kind of weird, right? It's such a...\", 'start': 114.286, 'end': 142.315}, {'text': \" I don't like that feeling there. Maybe it'll come back when they were, you know, it's like, well, there are these guys that just say anything and they just like, there's no whole bar. We could do that, I guess. But yeah, it's something's changed in the dynamics. Like a few years ago, I think nobody would have even thought about this censorship stuff. Like, you know, just say any crazy joke and that's fine. You know, like we're just having fun, but I don't know. Well, and there's, there are certain platforms that, that,\", 'start': 143.429, 'end': 172.437}, {'text': \" don't necessarily do that, right? Yes. But unfortunately, there aren't great alternatives to some of the bigger platforms. But the important thing is to try to reach as many people as possible with a message of the possibilities of what could happen and give people a perspective of what's really happening to your\", 'start': 173.551, 'end': 200.888}, {'text': \" I like to say to your currency, not necessarily your money, if we're talking about dollars of any type. Because there are still lots of people that don't understand that. But for those of us that are here to learn about charts or gold or stuff, we should drill into that a little bit more. So Patrick, you recently put a poll or a question up.\", 'start': 201.462, 'end': 231.128}, {'text': \" asking for the long term bearish chart for gold or the reasons that some of your Twitter followers are bearish on gold. So why don't you lay out the bearish case and let us know if anybody made a cogent argument against owning gold long term. Yes, I'll share my screen. Look, this\", 'start': 231.432, 'end': 256.863}, {'text': \" going to cover that. I'm telling you, a few charts later, people will understand what's going to be required macro-wise to have a bearish outlook on gold. Then you'll see, like I see it, what are the probabilities of having a bearish lookout on gold? Because you got to remember,\", 'start': 258.145, 'end': 278.041}, {'text': \" The price of gold going down is good. Okay, not if you bought it at a certain price and you wanna make money, like fiat money, right? Because you wanna buy an ounce of gold that you bought it, I don't know, 1300. Now it's 2000, you wanna cash in and that, but you gotta think about what gold actually is. Gold, let's say like, let's take the Warren Buffett definition of gold. It's just like a rock that produces nothing, does nothing. Okay, fine, it's a rock.\", 'start': 278.918, 'end': 303.927}, {'text': \" that rock, people are actually putting their fiat money to buy that rock. They've done it for hundreds of thousands of years. They've traded that rock for stuff. In the 1970s, it started at $40 unpegged, and now today, it's close to $2,000. If it's such a useless rock, why is it costing you more and more fiat to get that rock? That's the question people should try to answer. With charts,\", 'start': 304.096, 'end': 331.332}, {'text': \" I'm gonna show you how to answer that question. Why is gold going up in price if it's so useless? If it's just, you dig it out of the ground, spend a whole bunch of money to dig it out, and then you spend a whole bunch of money to put in a vault and then pay somebody to check that vault. It's like, when you think about it, it's kind of nonsense, but there's a reason why it's valuable. So right now I put this yearly gold chart, before we answer that next question there,\", 'start': 331.687, 'end': 359.007}, {'text': \" For gold. So here's the yearly closes, the wicks. What's fun about yearly candles is it removes a lot of the noise. Often people say, Oh, Patrick, uh, silver is not above all time highs, uh, since the 1980s, stuff like that. Look, blow off tops guys. It's human emotions, creating these parabolic run-ups that don't last very long. They're clickbait titles. Oh, silver reached $50 in 1980s. It reached $50 in 2011, but.\", 'start': 359.986, 'end': 386.176}, {'text': \" didn't stay there long. So it's clickbait. People compare peak to peak. But with yearly candles, it removes all that noise. It gives you more of a toned-down ebbs and flow of the gold price. I could have put a long-term moving average. It could have given you that same visual. So here in the 1980s, the peak of gold yearly close was $600. And then when the bear market\", 'start': 386.868, 'end': 412.028}, {'text': ' It broke down versus SPX, SPX broke out versus produced price index, a whole bunch of macro stuff. The yields had gone bonkers, gold was stretched from its moving average back in 1980. It did its 10-year bull run, it was done, and then it just hibernated. In 2005, I think this is five or six, five, it closed above its previous all-time highs', 'start': 412.535, 'end': 435.434}, {'text': \" on the yearly close. And after that, it just rocketed upwards. Look at that. Gold just went bonkers. You can't even see the GFC here. I keep telling people, find me the great gold that's so bad in the GFC. On the yearly basis close, in 2008, gold actually finished green.\", 'start': 436.666, 'end': 453.085}, {'text': \" So the green or it's a white candle, but that means it finished positive. So it started the year right here where my yellow mouse is. It went down during the GFC for a few months, three months in 2008, and it closed the year above the start. If I would have drawn a line chart, it would have just went up and up and up and up all the way to 2012. Then after that huge red candle trend line broke down. And now look in 2020,\", 'start': 453.085, 'end': 477.672}, {'text': \" I'm actually doing the bull case now, but we haven't talked about the bearish case, and I'll tell you right after. Here in 2020, we had a close, just like we had in 2005, above the 2012 highs. Solid, solid close. And for the past three years, we've just been consolidating above that neckline. So this neckline, for me, it used to be resistance. Now it's support. So for me, I see that chart. I did that chart, Tom, and then I asked myself, Jesus,\", 'start': 477.824, 'end': 505.921}, {'text': \" Am I the only one who sees this as potentially it's a bullish shut up? We're not doing new lows. This consolidation is not below a breakout line. It's not lower down here. So I'm thinking about all these people calling long-term bearish gold, and I forget who it is, but I put it out there. Give me some solid evidence, guys, some solid objective evidence that the gold chart\", 'start': 506.242, 'end': 530.12}, {'text': \" is looking down like the 1980s, right? That it's stretched from its moving averages that show me something guys, that's going to tell me that gold's going to go instead of going up here to 2,600 and above, show me something that's telling me that this is going down to a thousand. Okay. Or 600, show me that.\", 'start': 531.132, 'end': 549.34}, {'text': \" And I went through all the polls this morning. I saw nothing seriously. Look, I had this, the, the, the Warren Buffett quote. So that's why, Oh, that's another one. Yeah. Like he must have seen this, right. Okay. Absolutely. But the price is still going up. So.\", 'start': 549.712, 'end': 565.034}, {'text': \" That's not quite the answer why it's a bearish case, right? You need, and then the BRICS, people said if the BRICS doesn't kick in, then maybe the gold chart, like there were no BRICS in 2002, or there was no BRICS in 1974, in 1970, right? And gold still went up from 40 to 2000. So I didn't see- I think it's interesting too, Pat, it's something that you and I touched on again, before we started recording this morning, just looking at this longer term chart, right? You know, you made the perfect point.\", 'start': 566.097, 'end': 595.882}, {'text': \" talking about the great financial crisis, gold over the year ended up closing positive. Having that sense of perspective, especially when we're talking about something specific, is really important.\", 'start': 597.215, 'end': 614.41}, {'text': \" It's very important. Look, even March Madness there, where it was my 2020 candle, you can't even see it there. That's 2019. Yeah, one of these candles here, you can't even see what happened in 2020 when gold went down 15%, 20%. You can't even see it. I'll tell you now, maybe I'll show you one more chart. I have a bearish case for gold. The next chart,\", 'start': 615.862, 'end': 640.111}, {'text': \" will tell you what needs to happen for gold to have a bearish case. I'll do it interactively, guys. I'll put purchasing power, guys, right here. That's from the Fed.\", 'start': 641.107, 'end': 660.867}, {'text': \" Remove indicators. So that's from your purchasing power is your US dollar adjusted for inflation since, goodness, from 1913, guys, we have this data. So it's from the Fed. I'll show you what happens to gold. Well, gold was pegged. I'll put gold. I could put silver or oil, which is a little bit better because silver and oil, they were not pegged like gold prior to 1970. So that's the gold chart, bottom left to top right, as\", 'start': 661.036, 'end': 687.58}, {'text': \" Purchasing power goes from top left to bottom right. I'll put the silver chart, guys, because it was unpegged. New price scale. Here's my silver log chart. OK, here it is. You see here gold was pegged, and sometimes they readjusted it there. But if they wouldn't have pegged gold price, it would have looked practically like the silver chart. So I'll remove the gold chart, and then after that, we could circle back to it.\", 'start': 688.103, 'end': 714.58}, {'text': \" I'll inverse the purchasing power, guys, just to show you. Now, the purchasing power, the more it goes down, it looks like it's going up. Look how silver tracks it. The times that silver had huge drawdowns, if you remove these spikes here, it was when the purchasing power actually went up. The bearish case for me for gold is if they're able to actually\", 'start': 715.728, 'end': 741.833}, {'text': \" increase your purchasing power. So either the US dollar, or it's going to be a combination of that. So purchasing power is a US dollar adjusted for inflation. If the US dollar goes up and inflation goes up not as fast, your purchasing power is going to go up. If the US dollar goes down, but inflation goes down less, then your purchasing powers go up. That means your US dollar will be able to buy you more. And I'll show you the big, big picture for that purchasing power index.\", 'start': 744.668, 'end': 769.964}, {'text': \" This is a seven-year rate of change. So whenever you have a chart, Tom, and somebody shows you from bottom left to top right over long periods of time, your first reflex is, and all your viewers' first reflex is, okay, what's the rate of change? The market loves pricing in when something changes, like some type of behavior changes fast. Because while this thing is going down, and it's seemingly just going down, oh, here you could probably notice it's going down a little bit faster. If you put a rate of change indicator, you could actually spot\", 'start': 771.972, 'end': 801.335}, {'text': \" when there's an acceleration and destruction of purchasing power. And look at that here. Here is the top. Purchasing power plummets a seven-year rate of change. That's when silver went bonkers. Here, the seven-year rate of change topped out in 1964. Then it started rolling over. Guess what happened right here in 1967? Silver started going up. Then after that, we had the bull run for oil, gold, silver, kept going up here.\", 'start': 801.908, 'end': 829.702}, {'text': ' Here, look at that. While the purchasing power kept going down throughout the 80s, its rate of change was reflating. Here below zero, the seven-year rate of change is negative, but it was still reflating the rate of change. That gave', 'start': 830.174, 'end': 844.653}, {'text': \" headwinds for gold, silver, oil, and it gave tailwinds for growth stocks, stuff that likes cheap money and all that. But look what happened here in 2020. The rate of change just got destroyed for the purchasing power. And here, another trend line. They're starting to falter one by one, these more and more slanted trend lines. We're in the same type of angle here. I'll put an angle thick.\", 'start': 845.682, 'end': 875.534}, {'text': \" 51% here, 51 degrees. Look at that. We're at the same type of angles, even more acute than we were in the 70s. So for gold to start going down in price, you would need this to go up just like it did\", 'start': 875.922, 'end': 893.32}, {'text': \" the 1980s, right? That's what happened. So as soon as this bottoms and starts turning up, the rate of change for your purchasing power starts going up, then that's probably for me, that would be a bearish long-term evidence that gold is going to go sideways or down. You need something like that. You need the rate of change to go back above its seven-year rate, seven-year EMA moving average. It's not happening at all right now. And here just in the 2000s, gold's bull run here,\", 'start': 894.248, 'end': 924.218}, {'text': ' black thick line. Here, the 2004, the rate of change, just that small rate of change going down below its seven-year EMA and going all the way down here, that fueled oil, fueled silver, fueled gold until the rate of change right here back in 2011 started breaking out. And guess what happened in 2011-12?', 'start': 927.02, 'end': 948.62}, {'text': \" gold top, silver top, oil topped a little bit before, and that whole period here where the reflation in your purchasing power, it's still negative, you're still below the zero line, but because the rate of change was going up, that was enough headwinds. I'm taking a pause when I say headwinds because I'm dyslexic, I always flip the terms and I have to think it through. That was headwinds for gold, silver, and all these commodities, but the script has definitely flipped right now, definitely flipped.\", 'start': 948.721, 'end': 977.864}, {'text': \" That's the big picture. That for me is the bearish case. The macro bearish case for gold is purchasing power has to start reflating, Tom. Right. But I think considering all of the points that it looks like we're facing, these problems that we're facing, whether they're economic, whether it's stock market related,\", 'start': 978.016, 'end': 1005.758}, {'text': \" government spending, the idea that we can't keep servicing the payments on our debt. The balance sheet reduction that we have seen in the last year, I don't think that's a trend that continues. I could totally be wrong, but once they start printing more money again, I think we get a major reversal in that purchasing power chart again.\", 'start': 1006.602, 'end': 1036.285}, {'text': \" But going back to, let's say another point that you made earlier about the price of gold and that longer term perspective again, show us that chart, Patrick, of house prices, but priced in gold. Yes, yes. And this is another, I'm sure a lot of your viewers, they're enlightened to this. They understand why gold protects your purchasing power because\", 'start': 1038.057, 'end': 1066.61}, {'text': \" Everybody's pricing stuff in fiat. Because of that purchasing power, you saw that purchasing power chart practically over 100 years, top left to bottom right. So you know your purchasing power. It's the design of the system to keep destroying it. Like you said, they're going to start printing money again, and then it'll be more inflation. And then after that, that's going to erode the US dollar strength versus other currencies.\", 'start': 1067.369, 'end': 1088.699}, {'text': \" Look, this is a single-family home for the states in the American single-family home. Bottom left here, it was, I don't know, it was $18,000 or $24,000 for a single-family home. Today, $412,000. And people think they're rich. Oh, cool. My house went up in value.\", 'start': 1089.155, 'end': 1105.928}, {'text': \" Is it really your house that went up in value? Or is it your fiat? I made a funny voice. Is it your fiat that lost purchasing power? And you could answer that question. So while there's cycles, yes, of bull markets and bear markets in the housing market, of course, like the dynamics of supply and demand, they keep fluctuating. As soon as you price your housing pricing in gold, you'll see that it went sideways. It's been going sideways.\", 'start': 1107.464, 'end': 1136.32}, {'text': \" while oscillating over long periods of time, it's been going sideways. So today, it costs you about 213 ounces of gold to buy that family house. That line, if it could dry even to 1970s, it was practically the same amount of gold that was required to buy that single family house. Gold, so that means if you would,\", 'start': 1136.742, 'end': 1161.127}, {'text': \" held gold, or if you would have been paid in gold, in ounces of gold, your wages in gold would have never changed. Let's say you were paid 10 ounces of gold per year. Today, you'd still be paid 10 ounces of gold, and you'd still be able to buy the same house. That's what it's telling you. But because of that's not happening, the wages you had back in the 70s, whether it's $5 an hour or $1 an hour, and now it's $50, $100 an hour, that's what was required to buy that same house.\", 'start': 1163.371, 'end': 1193.02}, {'text': \" This removes completely the illusion of fiat time because now people have to realize, oh, I'm making more money. Well, are you really? Because you're still buying a house. You're still buying the same house that was back there, but now you need more of that fiat that's losing money. And I'll tell you what's sad about this is if everybody was paid in gold, everybody,\", 'start': 1193.999, 'end': 1213.709}, {'text': \" be able to purchase the same house. But now people are paid in fiat. If you're getting a pay raise, Tom, and I'm not getting a pay raise, now there's more fiat running after that same house, then I'm losing purchasing power. That's the wealth gap or something you're starting to wedge in between the ones who are able to get wage increases and the ones who cannot. They're just getting trapped, and they're getting separated, and this is going to hurt them. So\", 'start': 1214.114, 'end': 1243.223}, {'text': ' I could do another example I did with pricing crude oil.', 'start': 1243.645, 'end': 1248.067}, {'text': \" Same thing, crude oil, here's from the 1861s, guys. Crude oil price in fiat cycles within it, bottom left to top right. But look, price in silver, guys, in silver, it went sideways. Yes, there are cycles of up and down. But when you compare those two commodities, there's no silver dilution. There's no destruction. They aren't printing trillions of silver. I don't know what the annual production of silver is, but it's just enough.\", 'start': 1248.573, 'end': 1273.835}, {'text': \" to create this beautiful, oscillating, sideways channel that's been there for, God knows, like over a hundred years. It's been going sideways, this relationship between two commodities. That's the way it should look like. But as soon as you put in fiat, you play the fiat game. And if you're not playing the fiat game and you're not outpacing destruction of purchasing power, you're actually falling behind. That's like, when I say I put these charts out, I said, oh my goodness, man, hopefully,\", 'start': 1273.835, 'end': 1301.881}, {'text': \" And that's one thing about charts. Sometimes, you know, I do TA, like I show you all these cool metrics, but here I just have two arrows and hopefully that visual will help.\", 'start': 1302.944, 'end': 1314.554}, {'text': \" awaken people. This chart, I should be able to show to my father, and he should say, oh my goodness, you're right, Patrick. It's crystallizing now. This is what's been happening. I've been living in this rat race where I'm so happy my house went up from 400,000 to 800,000. Yes, maybe if you got your timing right on that buying and selling that house, you did great.\", 'start': 1315.398, 'end': 1336.88}, {'text': \" And you locked in that fiat. But as a society, as a whole, over hundreds of years, we've been losing. It's just we don't remember, right? Because I live, I do my rat race thing, I die, and then it's my kid's turn. So we have shorter term memory. But if we take all these generations of people, what the game plan is destruction of purchasing power. And I'm sure there's examples. Your guest must have brought that up. The Romans did this, like all the other civilizations, right? If they had charts,\", 'start': 1336.88, 'end': 1363.458}, {'text': \" of stuff priced in their fiat or their gold clippings, they cheated, they melted the gold, they put cheaper metals in it, and compared to a pure gold price, it would probably look like that. Yeah, exactly. And again, I think you make a good point there, Pat. And that's the idea that I go for too, is that why I have different voices on the show is to have\", 'start': 1363.458, 'end': 1390.205}, {'text': \" different ways of explaining fairly similar things. But sometimes, let's say, one guest will explain it in a way that you can understand it, and another guest won't explain it to you. But that doesn't, just because they didn't explain it to you that way doesn't mean it won't hit me a different way, right? We have example of kids at school, right? They're there. And you remember some teachers?\", 'start': 1390.795, 'end': 1417.745}, {'text': \" the subject matter, you're there and, oh my goodness, what is this? It's Chinese. I don't understand it. And then you get another teacher who just has this perfect analogy, this perfect example. And that's why some of your great guests, they use great analogies that people, ah, that's what's happening. Peter Schiff is great at that. There's a whole bunch of guys that are super good at that. And that's all you need that\", 'start': 1417.762, 'end': 1439.918}, {'text': \" that click moment, and then everything crystallizes. So we're kind of teachers, and that's what's great about you having a whole bunch of variety of guests. Even if you think the subject's repetitive, it still has value because the way it's presented,\", 'start': 1440.88, 'end': 1455.68}, {'text': \" if it's a, you're able to transmit the message to somebody who needed that different tangent to, to crystallize that information, what's really happening. So yeah, I mean, charts are useful for that. If we're able to create some type of visuals, because often people say, this is the thing. If you don't have charts, it's you hear stuff. It makes sense.\", 'start': 1456.743, 'end': 1478.545}, {'text': \" But then after that, it's hard to really grasp visually, like, is that really the truth? That guy, is he just a conspiracy theorist saying that governments have been depreciating our money over it? Oh, did I say a bad word? Is that where you're gonna get us a shadow ban? No, we're good.\", 'start': 1479.355, 'end': 1492.771}, {'text': \" But anyways, a chart shows you objective truth. So no matter what you think, no matter, and it helps also frame it in the context of time frame, right? So now we're talking about hundreds of years, guys, right? Because I know people say, oh, how did silver protect my purchasing power? It's down since 1980. Look, fine. And I kind of explained it. There's blow-off tops. There's all these shorter-term mechanics. But if you take a step back,\", 'start': 1493.75, 'end': 1518.472}, {'text': \" over long periods of time, the game plan is clear, clear as night and day there. It's destruction of purchasing power, save the US equities, US equities start going down, let's print some more money, destruction of purchasing power, keep everybody happy, but we're all frogs in the boiling pot of water, Tom. That's an analogy, and we're boiling.\", 'start': 1519.535, 'end': 1541.017}, {'text': \" And I think we're all going to die in it. Nobody's going to wake up in time, right? Or maybe people do wake up in time, they get out of the boiling water, they start understanding this, they start buying gold, or they start doing stuff to their portfolio to protect against this destruction of purchasing power. But I'm afraid most of the population is just stuck in that\", 'start': 1541.354, 'end': 1560.541}, {'text': \" that boiling pot there. I don't know how they're going to get out. The system's not built to get people out of that thing. It's to encourage, oh, I got a raise. I got more money. But they don't understand why getting more money is not something's wrong. I was trying to explain that to somebody. It's like,\", 'start': 1561.553, 'end': 1579.627}, {'text': \" shouldn't be getting wages. Maybe we should be getting wage increases, but not as fast as we're getting them. Prices of stuff should fluctuate within a range, but there's no reason my loaf of bread that cost me $1 in 2000 is costing me $7 today. It's like the bread's not different. What, special super bread? What happened? I think it's even the bread, you get less quantity for that same bread, right? So\", 'start': 1582.124, 'end': 1605.513}, {'text': \" you get that shrink inflation stuff. I can't even mention that, but if you met the shrink inflation and it's costing me $7 to buy that same thing of bread, it's not the bread that changed guys. It's like the gold thing. It's not the gold that changed. It's your Fiat that's getting destroyed. And no matter what reason you think why that's happening, it's, it's happening guys. So you gotta, you gotta prepare for that. So Pat changing the subject a little bit, how do you use,\", 'start': 1605.513, 'end': 1634.774}, {'text': \" what you call a heat map for looking at new themes to start playing out here and, and what goes into that? Yes. Yes. We could definitely cover that. Let me pop it up. All right, here it is. Okay. So here's the heat map guys. I have, there's 11 sectors in the SPX. So they're, they're listed down here, energy technology, industrials. And then after that I cross,\", 'start': 1634.825, 'end': 1660.812}, {'text': \" verify their performance versus the other sectors. And if it's a black spot here, it's because I'm comparing energy versus energy, right? So it's going to be a, it's going to be neutral.\", 'start': 1661.268, 'end': 1669.165}, {'text': \" So the top sector right now in the 11 sectors, and it's been like that for a little while is energy technology used to be at the bottom, but it had a great rally. So this, the most performing. After energy was technology on that crazy rally, right? And video, everything going bonkers, industrials, consumer dictionary, discretionaries, uh, uh, services, healthcare, materials, consumer, stables, real estate. It's like the bottom has been the bottom for a while.\", 'start': 1670.397, 'end': 1694.765}, {'text': \" So this tells me, and this is what's important guys, this is a huge advantage that the SPX has over, let's say Bitcoin or any instrument that is just an instrument like a company, right? If Enron goes bankrupt, Tom, Enron cannot do like the SPX. SPX is the 500 top companies. So that's why the SPX has survived a hundred of years. And it went from bottom left to top, right? It's able to always get rid of the less performance,\", 'start': 1696.638, 'end': 1722.338}, {'text': ' stocks and always keep the 500 top performing stocks. SPX has this huge advantage. But what happens when the SPX starts going down in a bear market,', 'start': 1722.625, 'end': 1733.375}, {'text': \" It's like a sector rotation happening. Energy was probably underweighted at the beginning, at the end of a bull run. Let's say in US equities, energy is probably like 5%, 8% of it, technology probably a higher percent. Then as technology gets sold off, or the ones that have been performing for 10 years in the SPX are going down, the sectors that have been underweight\", 'start': 1734.758, 'end': 1756.257}, {'text': \" They're going up in value, right? People are reallocating to purchasing power sensitive stocks, like I showed you the oil chart tracks inversely purchasing power. And after that, that helps keep a bottom, right? Because if the SPX was just tech stocks, the\", 'start': 1756.257, 'end': 1771.327}, {'text': \" whole SPX would look like, it could have looked like the ARK ETF, right? It'd just be like, it'd do a bell curve and go to zero. So that's a huge advantage of the SPX. But you could play, even if you're not, and I think we talked about this before, often people are calling the death of US equities, SPX to go minus 80%. Well,\", 'start': 1771.664, 'end': 1789.518}, {'text': \" kind of tough to do over long periods of time, because what's going to happen to the SPX is while some sectors get destroyed, and yes, it could create a vacuum downwards, other sectors are going to help keep it afloat until one day energy is going to be maybe 30% of the SPX, and then energy is going to be topped out. And it's going to be those cycles that repeat over and over. So people are always looking for some death scenarios for SPX. You got to be careful. It's able to rotate out the bad stuff. And I'm not even talking about dividends. There's also dividend paying companies in SPX.\", 'start': 1789.855, 'end': 1819.842}, {'text': \" So Bitcoin cannot do that. Enron could not do that. So you've got to be really careful when you're dealing with instruments that are not an index. They're not able to rotate out the bad stuff. If they go bankrupt, like let's say Canopy Growth, a wheat stock, it went to the moon, and then it did a bell curve. It went back down to zero, right? So be very, very mindful of that. So in the heat map right now, I know that the energy sector is kicking butt across the board, right? And it's tracking the yield.\", 'start': 1819.842, 'end': 1847.753}, {'text': \" And the average score while it's beating 10 of the 10 other competitive competitors sectors has an average score of 90 based on our proprietary way we score this. So I'm keeping a good, good track on this and any stock, any, any sector that used to have a below 50 rating Tom, and it goes up above 50, I'm going to get an eye on that. So you want to probably be look more closely at these charts. I'll just show you one, what it looks like. So here is,\", 'start': 1848.479, 'end': 1877.42}, {'text': \" technology versus energy. I could click on that, and I'll have the daily chart pop up with some annotations. Here is energy. This is what happened to the energy chart, guys. Energy bottomed versus technology in 2020. While technology has gone up,\", 'start': 1877.993, 'end': 1896.083}, {'text': \" recently. So this is technology outperforming energy. Now energy has bounced versus the energy sector has bounced on the 700 day moving average versus technology. And it's trying to go back in that trend. So right now the way to see it is technology and energy have practically gone sideways since, since February, 2020. So for over three years, both these sectors now energy,\", 'start': 1897.535, 'end': 1923.893}, {'text': \" drastically getting destroyed by technology, energy going back up, trying to halt that downwards momentum versus technology. Now, technology did a bear market rally. It's flat right now. If ever energy is able to bust above this remaining moving averages, and I'm probably going to have to redraw my line here, if ever it starts breaking upwards, and look, energy has been doing higher lows versus technology throughout all of this.\", 'start': 1924.45, 'end': 1953.543}, {'text': \" That should help you guys think, okay, sure. Nvidia, all that stuff went up, but look at, look at the sector, which is beating all the other sectors and potentially could be breaking out versus technology. That's probably where you want to look at to start putting your, your eggs, you know, whereas, okay, this is a, it's outperforming all the other ones. Cause this is a big question. People always have to ask themselves while you could be in an instrument Tom, that's going up.\", 'start': 1954.505, 'end': 1983.007}, {'text': \" are you really in the instrument that's going up fastest? Because if I gave you the choice and you're going at the racetrack, that horse is pretty fast. But there's three other horses that are going faster. Once you bet on the horses that are going faster, why would you bet for the guy that's going to finish in third place? So that's what it does. The heat maps, they help you compare sectors. So you're always able to\", 'start': 1983.614, 'end': 2006.193}, {'text': \" load up or be more heavy weighted on the sectors that are outperforming all the other ones. Until that trend reverses and after that, then you could flip into sectors that are gaining steam versus the one you're at. But right now,\", 'start': 2008.42, 'end': 2020.942}, {'text': \" Energy looks great. Uranium looks great. All those sectors, they've been leaders and they held up really well amidst the technology move that they did recently there. They just did great. So yeah, hopefully the heat map could help people. I look at it every day, Tom, and people should do that. Always compare, use racial charts, always compare versus other instruments and make sure you're in the fastest horse.\", 'start': 2022.207, 'end': 2048.482}, {'text': \" So Pat, maybe if we could, can we go through a uranium chart and show some of our listeners what you're seeing from your perspective that is making you excited about that sector? Holy Moses. Well, I'll go with Cameco because Cameco is, everybody knows about Cameco. I'll put, oh my goodness, that chart is just bonkers.\", 'start': 2050.388, 'end': 2075.515}, {'text': \" Cameco in the 1980s and then the 2000s is something to behold. Imagine being in this thing, and I know I'm going for the emotions here, but this is what Cameco did there. Went above, it's one year and three-year moving average, it's trying to garner good momentum, pulled back, and now it created one, two, third reaction, so now you have a valid breakout line.\", 'start': 2077.625, 'end': 2101.047}, {'text': \" I'll zoom in to show you guys. Here is a beautiful, not once in a lifetime opportunity, but you should look for these huge basing patterns breakout. And look at this round, Tom. It's like it never went below the one-year moving average. That's so powerful. Sometimes you see huge dips below the one-year moving average to hit the three-year here.\", 'start': 2102.397, 'end': 2121.719}, {'text': \" Month after month after month, Cameco just went up in its bull run. And here for two, three months went sideways. Now you're starting to have a topping pattern, momentum slowing down. You're starting to stall below the one-year moving average. That's how tops are built. Look what it's doing today. It did this. Cameco just went bonkers back here. Huge breakout candle in 2020.\", 'start': 2122.411, 'end': 2145.918}, {'text': ' above the one-year three-year moving average, it went up, up, up, slowing down momentum. Here, it was starting to get into trouble here, but it resolved upwards. It had a decision, it could have resolved downwards. The three-year moving average, the longer term was helping support the price. Now, it broke out in June of 2023, it had a crazy, crazy breakout. We had our members on that, but', 'start': 2146.036, 'end': 2174.302}, {'text': \" I think on Twitter, people, they should have caught on. Are people talking, like nobody's talking, I guess people are starting to talk about uranium. You must have a feel for it, Tom, there. Is it starting to get steam, uranium, in the psyche of people? Well, I think there's always, you know, a particular subset of people that feel like that. And that's why I like to try to get, again, different points of view on it.\", 'start': 2174.757, 'end': 2203.765}, {'text': ' and look at it from a non-emotional point of view, like from a chart, right? If you can see that it is continuing to get momentum, the price is continuing to make higher highs and higher lows, that seems to be moving in a bullish direction, right?', 'start': 2205.858, 'end': 2224.927}, {'text': \" Exactly. It's like the trend is your friend. I could just remove the price, guys, and I'll just hide it. Remove, hide, hide the price. Just look at the three-year moving average. Just look at that. Look at the three-year moving average, guys. That's\", 'start': 2225.585, 'end': 2240.705}, {'text': \" There's no narrative to attach to that. It's just going up. So for a moving average to go up, guys, it's basic, basic math. I don't know what year we learn about moving averages in school, but that means for a moving average to go up, the price has to be above it. If the price is not above the moving average, then the moving average will start accounting as time moves on, then the moving average goes sideways as the price goes below it. But when you have a beautiful three-year moving average like that, if you want to know if you're in a bull run,\", 'start': 2240.873, 'end': 2270.523}, {'text': \" Look, it's kind of clear night and day where we are about, right? I'll put the price back up. So yes, now there's resistance. While everybody might start talking about Cameco, I don't know, maybe soon, it's approaching a volume-defined wall, a small one here. So this is all the people who bought back here in 2007, 2011. I'm not quite sure that's the purpose of a base. Often, these bagholders are gone.\", 'start': 2270.962, 'end': 2295.363}, {'text': \" But there could be some, I'll draw a little roadmap to help people see what I could see. It's possible right now that we have some type of consolidation before this thing keeps going up, just like it did here. You see that monthly candle went red, went down. Moving average, not too stretched. Still, there's room for the moving average to keep climbing upwards. That's my FOMO.\", 'start': 2296.223, 'end': 2319.19}, {'text': \" That's what I call my FOMO. Look, when moving average is so stretched, that's when everybody starts talking about, I don't want to say the word there, but the ASS Uranium guys. Did you ever hear about this? It's like this thing going on Twitter, they're the, like Uranium guys, they're supposed to be like arrogant people. You know, like there's like a running meme that they're worse than the crypto laser guys, the Uranium bulls. I don't know. But,\", 'start': 2319.629, 'end': 2349.43}, {'text': \" There were probably a lot of uranium bulls back in March of 2022, but that's when it was topping because it was stretched. This is the stretch from the three-year moving average here. If ever we get stretched 104% above, I'll do here, I'll put 104, my measuring tool. If ever we get stretched 104% above the three-year moving average, I'll zoom in for you guys. Here's a three-year moving average.\", 'start': 2349.683, 'end': 2376.953}, {'text': ' so we can get stretched all the way up here.', 'start': 2378.978, 'end': 2381.476}, {'text': \" Before we reach that crazy uranium arrogant geyser, it could be all the way up here before we have a two-year consolidation happening. That's when you have blow-off tops, you need a huge reset, you need to let the moving average catch up to the price. We're not there yet at all. A little bit of resistance just up around the corner, but if ever we climb back above that, there's probably going to be another move much higher. The longer this takes,\", 'start': 2382.235, 'end': 2409.185}, {'text': \" I'll show you how the moving average works. So over time, a moving average is probably going to go higher and higher. So my distance from moving average scales up. So the longer it takes and lets the moving average catch up the price, the higher I'm able to stretch above it, creating, it's like a scalable target mechanism. I'm able to always think that higher up targets are more and more possible.\", 'start': 2409.961, 'end': 2436.59}, {'text': \" And that's it for uranium. It looks great. I could do, I could show you one crazy chart. Okay. Name me. Okay. Let's say what? Uranium versus Bitcoin or uranium versus NASDAQ. Let's go versus NASDAQ. Okay. Let's, let's not pick on the Bitcoin guys. That's fine. I just don't see what uranium and Bitcoin would have, you know, to,\", 'start': 2439.408, 'end': 2464.282}, {'text': \" to relate to each other? Well, yeah. Well, in my head, you could have interchanged Bitcoin and NASDAQ because of their charts, they're both. Look at this thing. So this circles back to, if you have a chart breaking out on its own chart merit, like I just showed you, and now you're beating a competitor, you're beating growth stocks, you're beating tech, right? Energy, uranium energy versus tech, right? Remember that sector, that heat map I showed you?\", 'start': 2464.518, 'end': 2492.412}, {'text': \" Look at this thing, man. This is bonkers. Cameco is right there. It's bull flagging versus NASDAQ. This is crazy performance. If that thing breaks out, that's a bull flag. We're going NASDAQ. I'm forcing you, Tom, to hold either NASDAQ or Cameco.\", 'start': 2493.188, 'end': 2522.635}, {'text': \" then this chart helps you answer that question. If Cameco is able to close above this dotted line versus the NASDAQ, you have to overweight Cameco instead of tech. You're above a three-year moving average, above the cloud, above an important neckline. You're back in 2001, for crying out loud, and that's when the bull era started for gold, silver, uranium, oil. That's what happened in 2001. Tech stocks went sideways or down,\", 'start': 2522.973, 'end': 2548.91}, {'text': \" the other stuff, the money flows started going back into the other performance. So while even, OK, let's do the case where NASDAQ does melt up. Well, if this chart breaks out and the tech stocks are going bonkers, it's telling me that, I'm sorry, I'm away from the mic, guys. Sorry about the sound. The sound, that's telling you that Cameco will go even more bonkers than tech stocks. This is crazy. Look,\", 'start': 2550.04, 'end': 2575.522}, {'text': \" This is a blow off top here. This is a blow off top. When you're 10-year run, seven-year run versus NASDAQ, you're stretched from moving average, you're starting to break down below important levels. That's a top. This is a bottom. This is a bottom breaking out. I want to be bearish,\", 'start': 2577.057, 'end': 2598.893}, {'text': \" there's no bearish case here at all. The top's somewhere up here there. The long-term top is somewhere here, Tom. That's what's happening right now. We're in an intermediate sideways move. If that breaks out, it's going to be stepping ladders across the way for Cameco, the flagship. After that, we could have looked at other uranium, but if Cameco is not able to do its thing,\", 'start': 2599.737, 'end': 2627.345}, {'text': \" than the other uranium sector doesn't have a chance there. Chemical is the best performing uranium stock for a long time. I don't know, man. Does that look bullish or bearish? Looks pretty bullish to me. Unless I spent it. That's another thing, guys. Beware TA, guys.\", 'start': 2627.648, 'end': 2649.704}, {'text': \" and I'm being honest about this, I'm able to spin a chart in both ways. I could select my timeframes. I think we talked about this. Sometimes people talk past each other. I could have shown you a bearish outlook for Cameco by changing the timeframe, right? I could have zoomed in. I could have taken a different timing, starting point. I could have played with the scale. So guys, be really, really careful. People could spin any narrative by, I have a bias, I have a narrative. I could spin the chart to try to fit it\", 'start': 2650.379, 'end': 2679.134}, {'text': \" I could like nudge you in to have a conclusion that it's bearish or bullish. So you gotta be really, really mindful. TA is not done equally, but if it's done properly, it's not astrology for men. I'm talking to you, Max. It's not astrology for men. It can give you an edge. It can, it just can. It's like, it's, it's been there forever. The Japanese have been using it for hundreds of years. There's a whole bunch of books. It's human emotions. And I don't want to run too much on that guys, but if it's done properly, it can give you an edge guys.\", 'start': 2680.45, 'end': 2709.29}, {'text': \" Sorry, that was my rant. That's all good, man. As we're talking about TA, Pat, do you take volume or let's say more precisely lack of volume going into a big announcement, for example, or event for any chart into consideration? Like how do you use volume as part of your TA? I'll tell you. Let's do a masterclass on volume. So let's do\", 'start': 2710.083, 'end': 2733.033}, {'text': \" Cameco, or I could do, okay, I'll do the uranium miners. There we go, we'll give one out for the... So this is the uranium miners on the monthly chart. Check out the volume here, guys. A lot of people say, oh, crazy volume. You know, you often hear the news. People say, oh, the crazy inflows in SLV. Look at the volume. Oh, it's record volume. And people, for some reason, if you don't know about charts or you never did volume analysis, you think that's bullish. Oh, crazy volume.\", 'start': 2734.906, 'end': 2764.522}, {'text': \" Look, if there's crazy volume, that means a lot of people are talking about it. If a lot of people are talking about it, they already bought or they're buying, there's not much sellers left. You're probably closer to a top than a bottom. And look at this. Look at this crazy volume. It was a top. It was a top. So the way to see volume is you have to see it as a FOMO, a sentiment indicator. Like I'll do it again. It repeats over and over.\", 'start': 2765.095, 'end': 2793.158}, {'text': ' price goes up, explodes out, volume goes up, small reset in volume, explosion in volume, price goes up. And after that, once you peak, once you get that crazy volume, then after that,', 'start': 2794.593, 'end': 2810.354}, {'text': \" That was the top. But what's beautiful about this now is the volume has totally reset. And that's why I think not a lot of people are talking about uranium. Maybe we are, like you said, the niche people, but nobody's talking about it. But look what happened to the price, Tom. It didn't do a new low there. It didn't implode. It went sideways above an inclining three moving average, right below a 10-year breakout line. It's like\", 'start': 2810.523, 'end': 2835.886}, {'text': \" That's very bullish there. This is fuel, got to see this as fuel. That's how I read volume. Yes, you want volume on the breakout. It looks great. You want an instigator. I'll zoom in here. I don't know if people could see this candle here. Look at the volume right below my cross thing, my crosshair. It has four times the average volume of the previous four or five weeks. That was breakout volume.\", 'start': 2836.088, 'end': 2861.317}, {'text': \" But there's a difference between breakout volume and topping volume. And I'll show you something else that also works the inverse for bottoms. So look at here, I put Bank of America, crazy volume here. That's called capitulation volume. Same thing. So crazy volume, people, oh, it's going down. People are crazy, crazy selling. Well, look, it was a bottom, huge bounce.\", 'start': 2861.468, 'end': 2887.068}, {'text': \" volume reset, here another huge breakdown, crazy volume here as the price plummets. But because it's happening below the three-year moving average, those are capitulation bottoms. And this is why the absence of volume right now in Bank of America is super scary because there's no capitulation yet. There's nothing yet. This is just\", 'start': 2887.928, 'end': 2911.57}, {'text': \" Nothing's happening. I say it's complacency, just going sideways. Like there's no increasing volume on price action going up. There's no increasing volume on price going down. It's just totally neutral. So look, it's vulnerable right now. It could break out, but because it's, I know I went straight to Bank of America and you didn't ask me, but it's just to show the opposite, you know, to show volume in bearish cases.\", 'start': 2912.836, 'end': 2939.971}, {'text': \" that right now, because it's below three moving average, it's below this long-term rising trend line, it's kind of scary right now for the banks, for the financial sector, right? And again, there's so much evidence. When the banks go down, look at that 2000 here, when the banks start rolling over, then it creates\", 'start': 2941.017, 'end': 2961.959}, {'text': \" an entry point for other asset classes to go in, right? And the destruction of purchasing power required to save the banks often is going to feed into gold, silver, uranium, all the other assets, which are just rocks, just simple rocks, stupid stuff. But because there's an acceleration in the destruction of purchasing, I hope this sounds going to be good because I was leaning back a few times. It's all good.\", 'start': 2963.191, 'end': 2990.866}, {'text': \" because of the acceleration and the destruction of purchasing power. It goes back to the first, first chart, guys, I showed you, you got to remember that chart. Purchasing power is always practically going downhill, but it's when it accelerates downhill, that's what shifts people. Oh my goodness. That's when the hedge funds have to rebalance. That's when the trillions of dollars have to redo their thing. The margins\", 'start': 2992.064, 'end': 3013.613}, {'text': \" collapse, that's when big stuff happens. And yeah, it's crazy. Every time we talk, even if we have, we talked about four months ago, it's still playing out. It's just, these are big, big things. You know what I think? Even when there'll be a breakout in gold, we'll have a chat. It broke out officially, but it's still going to be the slow process. Cause now we're talking about like years for this to play out. You know, it's, it's just, it's just a long process guys. It's not going to be over,\", 'start': 3013.867, 'end': 3043.263}, {'text': \" overnight. It's not going to end overnight, what's happening right now. These long-term things are slowly grinding into motion there. Right now, the evidence shows that it's still on the bull case there for gold, silver, uranium. Nothing's changed yet. It's just playing out. That's why I think it's important to have that, to get some of those longer-term charts and perspectives.\", 'start': 3044.748, 'end': 3070.263}, {'text': \" to look at those and use those as instruction. Because, again, it's kind of hard to see the forest for the trees sometimes. You might be in the middle of it, but it doesn't feel super exciting or crazy bullish at the time. That's right. Look, 10,000% gain in silver, guys, you'll never hear that. Forget that stuff. That stuff is clickbait. Let that for the clickbait stuff. What you want,\", 'start': 3071.866, 'end': 3099.98}, {'text': \" The only way to make money guys, forget the percentages, the only way to make money is to identify a trend that's going up, get in.\", 'start': 3100.199, 'end': 3106.898}, {'text': \" and ride the trend. It's so simple, guys, so simple. Don't need fancy stories. You need that three-year moving average turning up like I showed you. That's all you need, because you can have the craziest bullish narratives for anything, your pog sticks, whatever you're playing. It doesn't matter. If the price is below that three-year moving average, get out of there. Get out. Can't fight those trends, guys.\", 'start': 3107.86, 'end': 3133.898}, {'text': \" price above an inclining moving average telling you the price is going up. That's how you make money, guys. How come you're not a billionaire, Patrick? Patrick, how come you're not a billionaire? I get that all the time. It's bonkers. Oh, man. I think you're in this process and just trying to learn as well as everybody else is, right? Well, look, I'll answer that question because I answered that billionaire question\", 'start': 3133.898, 'end': 3162.603}, {'text': ' The people who say that is because they do not understand risk and money management. Because if somebody is telling you the story, the clickbait story of the Lamborghini and all that, oh, $10,000 became a billion, that means they bet everything they had on that play to go there. What if that play', 'start': 3162.788, 'end': 3185.097}, {'text': \" didn't pan out. Do you think Michael Jordan always takes three-pointer shots? What are the high probability shots in basketball? They're the dunks. They're going to try to dunk if they can, because it's a high probability shot. If you want to build a successful career, a 10-year career, and have a whole bunch of points, you're going to do dunks, and you're going to do just regular two-point shots. You're not going to do just three-point shots. They're low probability. They're going to hit the rim. They're going to go anywhere. Somebody's spinning you the story that they got rich on Bitcoin because they bought\", 'start': 3185.603, 'end': 3211.608}, {'text': \" whatever, it's somebody taking a long shot, they bet everything they had, and if they would have missed that shot, they would have been wiped out. They would have been wiped out. So that's part A of why anybody\", 'start': 3212.502, 'end': 3225.395}, {'text': \" is not a billionaire using charts or anybody even using fundamentals, because they're risking money management, and they cannot bet. They cannot risk. I will never risk more than 2% per trade. So let's say I have a trade, Cameco, I think it's going to the moon. Cameco goes up 100%. My NAV is not going to go up 100%. My NAV is going to go up 5%, 6%, because I just bet a small amount of my NAV, guys. If somebody is selling you another story than that, they're clickbaiting you, and it's dangerous. And the second\", 'start': 3225.445, 'end': 3253.66}, {'text': \" Part of that is your starting capital. Because once you know that, let's say a financial trader makes 30 or 40% per year. Let's say I started with 10,000. You don't know my background. I started with 10,000, 10 years ago. $10,000 at 30% a year. Let's say I'm super good. I'm 30% a year and I compound it. My $10,000 after 10 years, Tom, is just 137,000. I didn't remove taxes. I didn't remove money that I would have needed to pay my living expenses.\", 'start': 3254.251, 'end': 3281.656}, {'text': \" So there's no quick, get quick, rich scheme, guys. It's just not. And if somebody, if you want to do this for the long term,\", 'start': 3281.656, 'end': 3289.807}, {'text': \" Risk and money management, and it's all about starting NAV. Because I'll tell you something, most new billionaires during the year, they probably started the year with $800 million in their bank account. Most new millionaires, they probably started with $700,000 in their bank account. They started the year, and at the end of the year, they're at a million. They didn't start at $10,000. Those are low probability events. And that's my beef I have on Twitter is people selling you\", 'start': 3290.212, 'end': 3316.655}, {'text': \" telling you that you'll put everything you have in that stock, you'll be a millionaire. What the hell are they talking about? How come they're not liable? That's something. But guys, please, risk and money management, and we all have different NAVs. And like you said, it's a work in progress. If I started off $100K, maybe I have $1.3 million today. You guys don't know. Nobody knows. And $1.3 million, guys, it's not that much. Seriously.\", 'start': 3317.363, 'end': 3338.508}, {'text': \" I'm surprised by how much a million dollars is not that much anymore. Like we're practically all millionaires now. You have a house, a car, like a million dollars is not what it used to be, right? When we were young, I remember when I was in grade six, Tom, it was crazy. It was in my grade six, I don't know, it was in 1980, I'm 47, I don't know, 12 years old, like 35 years ago, in my school yearbook, they put ambitions, I put millionaire.\", 'start': 3338.508, 'end': 3367.465}, {'text': \" But back then, millionaire was a huge number. Nobody would have dreamt of being a billionaire back in 1987 or 88, right? But now today, it's like, you're depreciating your money so much. It's like, so yeah, I went on a tangent, but guys always consider that starting NAV and risk and money management. Warren Buffett does it, pension funds do it. Why don't these pension funds do thousands of percents? They don't because they use proper risk and money management and they don't bet everything on any single horse.\", 'start': 3369.187, 'end': 3399.157}, {'text': \" And yeah, that was a good question, because people could, they'll stop them from going to all these tourist traps, these penny stock, la la land websites, where it's just, it's just total chaos, man. It's just like,\", 'start': 3399.933, 'end': 3416.842}, {'text': \" danger, danger, danger, danger, and bet everything. You want the Lambo. You'll never see me like showing off a whole bunch of this stuff. Cause it's, it's nonsense. And it won't make you happy anyways, guys. It's like, seriously, why you're judging me because I'm not riding a Lambo. Maybe I'm happy with my Toyota Corolla. Like I don't need more. It's like people have, their values are all skewed up there. They're all messed up there. It's like,\", 'start': 3417.972, 'end': 3439.083}, {'text': \" and the rich people will tell you this, their money does not make you happy. You just need enough money to take care of the bills, take care of your family. But other than that, it's just extra guys. It's just stare in a bank account. And you know, it's like, I don't know, man, for me, it's not that important. You know, it's just, I want to give out value, help people get out of that rat race thing. But\", 'start': 3440.078, 'end': 3462.775}, {'text': \" Striving for trillions of dollars is not the endgame. If it happens, it happens. But that shouldn't be your endgame, guys. It should be the process of being happy every single day, guys. Well, Pat, I think that's a good place to kind of wrap up on the value that you've given us today.\", 'start': 3463.012, 'end': 3481.355}, {'text': \" unless you have anything else that you'd like to share with us? No, Tom, it's like, uh, let's, let's catch up again. Hopefully it gives a, this gives value to your, to your viewers, different tangent, a different viewpoint and, uh, share, I'll share everything. You should get a lot more views. I know maybe you're not trying for that, but the content here is stuff that's an alternative to what you'll never hear on CNBC. You'll never hear that anywhere else.\", 'start': 3481.523, 'end': 3506.347}, {'text': \" And it's weird because it's stuff you won't hear elsewhere, but it's stuff that appeals to 80% of the people. All the stuff we say makes sense to 80% of the people, but it's not stuff you hear 80% of the time. It's so weird. Absolutely. Yeah. Of course, Pat on Twitter at BadCharts1 and NorthStarBadCharts.com, right? That's right. That's right. Thanks so much for your time today, Pat. Appreciate it. Thanks, Tom. Thanks for your time.\", 'start': 3506.347, 'end': 3532.992}]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "INFO:pytorch_lightning.utilities.migration.utils:Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v2.0.9. To apply the upgrade to your files permanently, run `python -m pytorch_lightning.utilities.upgrade_checkpoint --file ../root/.cache/torch/pyannote/models--pyannote--segmentation/snapshots/c4c8ceafcbb3a7a280c2d357aee9fbc9b0be7f9b/pytorch_model.bin`\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model was trained with pyannote.audio 0.0.1, yours is 2.1.1. Bad things might happen unless you revert pyannote.audio to 0.x.\n",
            "Model was trained with torch 1.10.0+cu102, yours is 2.0.1+cu118. Bad things might happen unless you revert torch to 1.x.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_672fd87a-aaad-4db3-86c6-1e230e8fe67b\", \"buS-n3wh6vg.json\", 4130782)"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Detected language: en (0.99) in first 30s of audio...\n",
            "[{'text': \" Happy to welcome to Forward Guidance, Andrew Beer, managing member of Dynamic Beta Investments. Andrew, good to have you on Forward Guidance. Dynamic Beta, why beta? Isn't everyone trying to pursue alpha in this world? Why are you focusing on beta?\", 'start': 4.733, 'end': 20.444}, {'text': \" It's a good question. So actually, we just changed our name to DBI officially as of last week. I think people often talk about alpha and beta being separate. Beta is the cheap, easy thing that anybody can get, whereas alpha is this esoteric thing. The reality is they're interconnected. If you buy equities at the right time or you buy bonds at the right time, you can generate an enormous amount of alpha through those kinds of\", 'start': 20.444, 'end': 46.718}, {'text': \" regular decisions over time. So for us, dynamic beta, what it really meant was that if you want to understand how hedge funds are making money, um, then the thing you want to understand what their betas exposures are because they change over time. So the guys who, you know, who liked value stocks before might like tech stocks. Now they might like emerging market stocks later. It's really trying to understand that. And so our business is really built around the idea of you, you want to find the right betas at the right time and invest in them. And then we have a particular way of approaching that.\", 'start': 46.887, 'end': 73.195}, {'text': \" Right. So what you do at DBI is pretty complicated, but my understanding is it's liquid alternatives and you're taking what hedge funds do and delivering it, replicating it with a much lower fee structure. So my first question is, did I get that right? And the second question is, how do you do that? Because these hedge fund people, they're supposed to, they went to the right schools. They're all smart. They've all got Bloomberg terminals. So how can you, that's at least the narrative. So how do you replicate and beat them? And you beat them by charging much less.\", 'start': 74.613, 'end': 103.908}, {'text': \" So we do something called hedge fund replication, which is it's actually a really simple idea, which is that if you can identify what these hedge funds are doing, but you can do the same thing that they're doing, but do it more cheaply, you do better over time.\", 'start': 103.958, 'end': 116.497}, {'text': \" You get the same quality of what they're doing. And then what we try to do is then also deliver it in ETFs. In Europe, we deliver it in the versions of mutual funds. And so as a result, what you want is people to, you want to take the good things that hedge funds are doing, but then simply repackage it in a simple and straightforward, easy to access way. And so this only works in certain circumstances. There are a lot of things that hedge funds do that are truly esoteric.\", 'start': 117.087, 'end': 145.319}, {'text': \" But I've been in the hedge fund industry for nearly 30 years. It's not that mysterious. A lot of people try to frame it as mysterious because it's great for justifying high fees and things. But you talk to actual hedge funds about what they do. And a lot of times, it's pretty damn simple. So what is it? The great advantage hedge funds have is that they are\", 'start': 145.319, 'end': 168.438}, {'text': \" they're less constrained, that they can change their minds. And so the really fascinating thing about the asset management industry is that the whole industry is built to move slowly. If you go talk to your financial advisor, and your financial advisor is going to give you a 10-year plan, and we're going to own this much S&P and this much these things, and as though anybody has any clue what these things are going to do over the next 10 years. But the whole ethos of it is we're the steady hand at the wheel.\", 'start': 168.64, 'end': 198.627}, {'text': ' And so you get to the end of 2021, and the steady hands at the wheel, all were betting that rates would stay low forever. And so, boom, inflation starts to come back. What did they do? Nothing at first.', 'start': 198.813, 'end': 212.65}, {'text': \" At first, they have to say, oh, no, no, no. It's ephemeral. It's the ports in LA. It's this. It's that. It's all these different things. And so people didn't believe inflation was coming back because they needed to believe it wasn't coming back because they told their clients the year before it wasn't coming back, and they're not supposed to change their mind every year. Hedge funds can. And so a lot of hedge funds in early 2022 said, oh, this is getting real.\", 'start': 213.292, 'end': 240.005}, {'text': \" And so so they did some things. They said, you know, we're not going to own bonds anymore. We're going to short treasuries and bet that rates are going up. You know, the US is going to keep raising rates. That's going to have all sorts of impacts in the currency and other markets. So so so the alpha, you know, this thing that we all want that comes from active management is often because people change their mind. And and the thing is, most of us, most of the asset management industry in general, like you think about the pension plan down the street.\", 'start': 240.68, 'end': 268.793}, {'text': \" You know, they don't change their mind every quarter in their portfolios. They want to put things in and not change it for the next 10 years. But about 20% of the time, the world changes a lot faster. And those are the times like 2022, where you really need certain hedge fund strategies that can adapt to those market conditions.\", 'start': 269.857, 'end': 286.867}, {'text': \" So I am familiar with the chase, the horse that just won the last race type of strategy that's familiar on financial networks. It's familiar with financial strategists. And I'm sure it's familiar with asset allocators. And the perfect example is\", 'start': 287.626, 'end': 304.535}, {'text': \" Not a lot of people were talking about floating rate strategies. I mean, I'm sure they were, but not a lot in 2021. And now everyone wants floating rate strategies, but it's like, guys, you missed it. Rates are at 5.5%. The time to get into floating rate strategies was two years ago. And, oh, growth has been outperforming. Everyone is always chasing performance, I suppose. One narrative is Peter Lynch, fantastic investor, mutual fund manager, so much of the money went into\", 'start': 304.67, 'end': 329.965}, {'text': \" Peter Lynch's fund right after they just had a phenomenal year. And so actually, the realized returns of the managed dollar was much less than the returns of the fund if it didn't have those hot inflows and outflows. There are a lot of anecdotes of\", 'start': 329.965, 'end': 347.093}, {'text': \" folks who made over 100% last year, or 30%, 40% last year, shorting treasuries, going long oil, all sorts of strategies that are, you know, there's a lot of volatility last year, and a lot of room for outperformance just because the range of outcomes was so wide.\", 'start': 347.093, 'end': 362.821}, {'text': \" But have hedge fund managers, and it matters how you define the hedge fund community, have they, we know they have underperformed the market, including their very high fees, but have they outperformed the market without fees at all? Because even though you could say, oh, yes, these hedge fund managers, they're so prescient, they went out of large cap stocks and they did these strategies, haven't a lot of those alternative strategies underperformed over the past decade when the real thing to do was just go to sleep?\", 'start': 362.821, 'end': 390.53}, {'text': \" Oh, absolutely. Right. And that's, and I, I mean, so, I mean, the genius trades, right. If you want it to be one of the best performing hedge funds out there, you would have loaded up on NASDAQ in 2010 and gone home and charged people 20% a year for that, which frankly kind of, I mean, you look actually a lot of hedge funds kind of actually look like that. And those guys, you know, are now multi-billionaires. The issue of changing performance, it's, it's just, it's, it's wired into us.\", 'start': 390.53, 'end': 415.083}, {'text': \" Right. It's it's exciting. Right. You feel like you've found because because, again, the whole I think I think if you're thinking about. Investing, right, investing is not something that is that's that's cold and objective, that that happens on paper objective, it's a it's a very emotional experience. And and you talk to your financial advisor and your financial advisor is is trying to assure you that I can help you grow your money.\", 'start': 415.775, 'end': 445.424}, {'text': \" between now and when you die or when you retire with the least stress amount along, along the way. You know, we've got this chart that we show where we say, Hey, we've got to, you know, we love this particular hedge fund strategy. It can statistically have a big impact on your portfolio. And I show kind of these charts, these kinds of statistically inclined charts on it. And then I say, I will tell you that there's not a single client in America\", 'start': 445.525, 'end': 470.703}, {'text': \" who hugged their advisor after 20 years for raising their Sharpe ratio. Right. And so, so when you're thinking about investment opportunities and you're thinking about, you know, why people make the decisions they've been, it's, it's, yes, there's an economic side to it, but it's the other side. That's really powerful. And so what happened, if you just think about the past few years in early 2022,\", 'start': 471.8, 'end': 496.353}, {'text': \" People couldn't believe that inflation was in because it required them to go to their clients and say, Hey, that stuff I told you six months ago was wrong. I'm changing my mind. Remember, I was only putting you into tech stocks because those were going to be the best opportunities. I've got all these fang stocks. Uh, they'll never go down when interest rates are zero. Uh, I'm buying a 10 year AA rated bonds for you with a one and a half percent yield. And you should be happy about it.\", 'start': 497.753, 'end': 521.007}, {'text': \" Right. And so, so, so six months later, that's looking really scary for people. I criticize it, but I think, I think one of the best things that's been done on the asset management side and the wealth management side is actually the growth of these stable model portfolios. Um, that, that, you know, because, because they are, they're designed to help that person say,\", 'start': 521.581, 'end': 541.443}, {'text': \" I've got a 5% allocation emerging markets not doing well this month. They shouldn't be tinkering in it and selling it and buying it or moving in and out of funds, et cetera. So it's a really, really positive force about 80% of the time. But 20% of the time, they're pointing in the same direction when everything's gone the wrong way. And that's what happened last year. Stocks and bonds both went down. And the most interesting thing right now\", 'start': 542.658, 'end': 568.477}, {'text': \" is that we've gone through this, everybody, you and I and everybody else has grown up in this world where stocks and bonds hedge each other. One goes up this year, then the other one goes down. In 2000 through 2002, equities go down 50%. Bonds soar during that period of time, and you get a smoother ride. In 2008, equities get crushed again. Bonds don't really go up a lot, but they don't go down a lot either. Last year, they both go down.\", 'start': 568.578, 'end': 597.434}, {'text': \" Yeah. So what do you do when you're sitting in front of a client and say, how do I give you a smoother ride? And the old playbook is now a scary roller coaster. And clients don't want a scary roller coaster. But what do you do when all of your tools, all of your diversification tools that used to kind of offset each other now seem to be moving up and down? They all went up in January. They all went down in February.\", 'start': 597.89, 'end': 623.067}, {'text': \" They all went up last month. Maybe they'll go down or they all went down last month. Maybe they'll go up again this month. And so so there's a it's there's an existential crisis on the wealth management side in terms of how do you plan for the next 10 years? And people aren't talking about it nearly as much as they as they should, because I think people were basically, you know, I think one of the one of the when the world changes fast, everyone hopes it's not going to stick.\", 'start': 623.168, 'end': 648.076}, {'text': \" Right. And so even in the beginning of this year, it's like, Oh, last year was awful. Look what happens to rates. Oh, but don't worry. By the end of this year, we'll be back to the old party. And now we're not. And so as time goes on, it just, it's, it's sinking into people that, wow, something really, really big has changed. And this is going to reverberate through portfolios, through markets, through all these different things. And, and, and, and I think that's the most fascinating intellectual question today. If you're an advisor.\", 'start': 649.443, 'end': 677.624}, {'text': \" So your firm is delivering hedge fund returns at a much cheaper price. I get that. So my question is not really so much to challenge that. It's why hedge funds at all? Hasn't their track record over the past 10 years not been great? Can you go over the track record of hedge funds since 2008? Because, I mean, is it correct to say it has been less than stellar? Oh, absolutely. I mean, it's basically, look, the hedge fund industry has gone through a very\", 'start': 677.624, 'end': 707.206}, {'text': \" Um, uh, it's gone through a process of institutionalization. Um, but the fees that people pay, you shoot yourself in the foot, right? There, there is a, so we'll walk us through the math on that. I wrote a, uh, an editorial for the financial times several years ago, 80% of the value creation goes to them, not you. You take a hundred percent of the risk and get 20% of the benefits.\", 'start': 707.206, 'end': 735.826}, {'text': ' and 80% of it goes to them. Why? Why do that? Right. And so, so there are if you but if you look at 100 of these guys, you can always find the guy who looks unbeatable today. The problem is that guy is your embarrassing, awkward conversation in six months or a year.', 'start': 736.147, 'end': 757.156}, {'text': \" And so, and so, but, but it persists because a lot of people are in that business, right? They tell their clients, we're going to find the best guy. What they do is they find the guy who was the best. And back to your point about, about Peter Lynch and Magellan, you can say, make the same analysis with Arc. You can do all sorts of different things. So, so take, so take a step back. So we, we, we're not product sales guys, right? Rather what we do is we say, H1s do a lot of different things.\", 'start': 757.375, 'end': 785.168}, {'text': \" Let's boil it down into, first of all, what do they do that's valuable? And there's some things they do that are really valuable that we can't copy, but there are a couple of things they do that are really valuable that we can copy. And let's do that. And so we manage an ETF in the U S that was up 23% last year. Not because of us, you know, not, not our great picks in our calls, but because the underlying managers.\", 'start': 785.742, 'end': 813.755}, {'text': \" got inflation right. And so they were up 20, which was amazing. We were up 23 because we were cheaper, because they're up 20 was really up 26, let's say, right. And then they took 600 basis points of fees, and they all went and bought, you know,\", 'start': 814.75, 'end': 830.09}, {'text': \" yachts from Russian oligarchs who were selling their yachts or something. And, and, and, uh, and we went up 23, we didn't do quite as well as they did before fees, but, but we, um, uh, but, but again, you know, we're charging a quarter of the fees, six of the fees that these guys were charging last year. And so our view is that, you know, if you want to be in the space and you want to be happy in five years or 10 years, you got to pick a strategy that's going to really\", 'start': 830.123, 'end': 856.533}, {'text': \" It's going to actually have some demonstrable benefit as opposed to something that just looks like a disaster relative to stocks and bonds in the 2010s. So you got to really focus on that and then you got to find the right way to invest.\", 'start': 857.41, 'end': 870.269}, {'text': \" So that 26% performance of the long short basket, you're saying they charge 600 basis points of fees, and that's 2% management, or maybe a little bit less management fees, and then the carry of 20% or something like that of the carry. So that's why it was 600 basis points, because when they do well, they get that. Yeah. Yeah. So this is a strategy called managed futures, or some people call it as CTAs. And basically, it's the most valuable diversifier\", 'start': 870.843, 'end': 895.852}, {'text': \" Uh, that you can, that we think you can put into portfolio if you care about liquidity and stuff like that. So just the, the rough characteristics of the strategy, guys who've been doing this for a living over the past 20 years have had no correlation to stocks and bonds. They've had returns of between stocks and bonds, not quite as good as stocks, but better than bonds. And they've had, it hasn't been, and they, they went up during two, during the dot-com crisis. They went up a lot during the GFC and they went up a lot last year.\", 'start': 896.797, 'end': 924.067}, {'text': \" So it is, from a diversification perspective, you just say, I'm going to try to construct the right portfolio for the next 10 years. Particularly, everyone's got two stools, sorry, two legs of the stool. And by the way, this is a credit to Eric McArdle from Simplify for coming up with this metaphor. He's a competitor, but we're kindred spirits in liking the space and trying to find a way to explain it. The problem is you don't have, you haven't had a good third leg to the stool.\", 'start': 924.927, 'end': 949.598}, {'text': \" When you have real estate, but then real estate is going up with stocks and bonds, when you've got this, when everything's going up and down together, you effectively have just...\", 'start': 949.598, 'end': 958.525}, {'text': \" You know, you're buttressing the existing legs of the stool. You're not adding a third leg. Managed Futures is really a third leg. So it does give you a better chance of giving your clients a stable ride, not a scary roller coaster, but a smoother ride because it kicks in when you need it the most. But it's also a very challenging space to invest in and to describe, which is why we came at it from a very different perspective.\", 'start': 958.778, 'end': 985.087}, {'text': \" And so your ETF that was up 23%, that was long, short, or managed futures, or are they the same thing? No, so it's just managed futures. Managed futures, I just wrote a paper on this in an institutional investor. And\", 'start': 985.087, 'end': 1000.645}, {'text': \" these guys build wave detectors, right? They are a much smarter modern version of those chartists who would say this thing is breaking out and therefore it's going to keep going up. So they hopefully apply a bit more math to it and try to come up with, but what they do is they build these wave detectors, right? And so what happens is when prices move, I mentioned this very human response, right?\", 'start': 1000.915, 'end': 1025.941}, {'text': \" So you bought a stock yesterday at a hundred, let's say you bought crude oil, right? And you bought it at 75. You bought a, you know, an ETF that does it. And three weeks later, it's at 85, right? A lot of people are going to sell it cutback. When I was managing portfolios, it's so hard not to take chips off the table. Cause I just bought it lower from here and I can totally see it coming back, going back to that level.\", 'start': 1025.941, 'end': 1051.878}, {'text': \" So what these guys do is they basically kind of scan the markets and they're looking for these kind of the beginning of waves. And so when interest rates start going up, they detect it early and they jump on it. When the dollar started to get strong, they jump on it. When equity started to go down, they jump on it. And so they're not first into the trade.\", 'start': 1052.468, 'end': 1074.018}, {'text': \" but they tend to ride it a lot longer than a normal human being can ride it. So it's a little bit different from equity long short. The reason it's called futures is because if you want to bet that crude oil prices are going up, it's a really bad idea to go buy barrels of crude oil and put them in your garage. So go just, you buy a futures contract, you buy an ETF. But the way these guys are set up is basically, you know, if you want to bet that oil is going up or you want to bet it's going down, right? Waves go up or down, then this is the way that we implement it.\", 'start': 1074.912, 'end': 1103.92}, {'text': \" And so in 2015, we basically said, this is the best damn strategy we've seen from a diversification perspective. And it's one of the worst strategies to actually try to get to make into a good investment for a variety of reasons. So let's try to build a better way to do it.\", 'start': 1104.123, 'end': 1117.707}, {'text': \" So Andrew, you know, I feel like I began to half understand what managed futures are, like the 40th time that I heard the term. So I feel like for folks who are hearing it for the first time, and there are definitely our folks listening to this, yeah, we need to catch them up. So basically, stocks and bonds, other things that correlate commodities, gold, but when people say commodities, a lot of that is long only commodities. So what you're going to buy some sort of basket of wheat and copper and\", 'start': 1117.876, 'end': 1141.147}, {'text': \" A lot of time, you're paying roll cost if it's in Contango. And who knows? It's dependent on the regime. Managed futures is going long, cotton, going short, coffee, all sorts of stuff. And it's just mathematical and computerized. And it kind of sounds too good to be true. But I know that the back test and looking at the history, it is quite good in terms of not just the performance, which are less than stocks, but when you combine them with\", 'start': 1141.147, 'end': 1164.113}, {'text': \" stocks and bonds, it actually has a very good statistically shown benefit, right? So how is that kind of possible? Where is the alpha, you know? Right, so it goes back to this idea that this business is designed to move slowly and it's run by human beings, right? And I would just make a point about Managed Futures. The guys who've gone into the Managed Futures space are technical and they want to talk to guys who are technical.\", 'start': 1164.113, 'end': 1193.307}, {'text': \" Right. It is, it is, they are engineers by training. And so they want to describe things like contango and, you know, roles and this and that other things. I find that totally uninteresting from an investment perspective, because, because to me, it's about, as you say, why should this thing do well over time and why, and, and, and, and, and how is it good for me? Right.\", 'start': 1194.218, 'end': 1217.152}, {'text': \" And in a sense, but the people in the space, this whole space has been on a 30-year path toward commoditization, right? These guys are using computers to figure out whether futures contracts are going up or down. Who knew how to do that in the early 1990s? You know, like when I started at this firm called Dow Post, I mean, you couldn't get, it's really hard to get good data. You had Bloomberg, but even getting the, so it was this esoteric area where some people went into it. Today, it's pretty commoditized.\", 'start': 1217.624, 'end': 1246.632}, {'text': \" But the fee structure, like I mentioned about hedge funds, we're still a lot more if it goes to them than we wanted when we were looking at investing in the space. But the reason it generates its value additive to your portfolio over time is because it's going to react in a way that your financial advisor is not going to react.\", 'start': 1246.801, 'end': 1269.363}, {'text': \" Your financial advisor, first of all, is not going to bet on crude oil collapsing 20% over the next six months, right? So that's, it's something they can bet on that. And it's making money in areas that you wouldn't otherwise invest in yourself. And, but what they really make the big money is when the world changes faster than advisors, model allocators, et cetera, can react. Because it's really hard to go back to your client\", 'start': 1270.139, 'end': 1299.907}, {'text': \" and and tell them that you think the world may have changed and you're going to make a wholesale change in your portfolio. So, you know, so I go back to you, Jack, and I say I say, OK, so I know last December we told you and we showed you data that showed that 60 percent of economists in October in September thought that there would be no rate hikes next year. You know, I know we told you is transitory. We know told you this. It's going to be disrupted away. There's never inflation is ever coming back. Well,\", 'start': 1299.974, 'end': 1328.476}, {'text': \" We've totally changed our mind. We're selling all of our bonds. We're selling our growth stocks. We're buying volume stocks. We're shorting bonds. We're doing all this other stuff. You're going to be like, god, that's really, really, really interesting. I'd like my money back now, please. Because you've lost your mind, right? You're not the steady hand of the wheel anymore. I'm supposed to be the one. Or as I should say, you as the client are supposed to be the one who panics about this stuff. Your advisor is supposed to be the one, the steady hand through the seas.\", 'start': 1329.725, 'end': 1359.205}, {'text': \" And now let's say, let's say the advisor does change their mind and they're wrong. You had a great business. You just blew it up, right? You've got clients who, if everybody goes down like last year and you go down, your clients aren't happy, but you're okay. They don't expect you to be a hero, but imagine if everything goes up and you're the guy who's down, you're looking for a new job. And so, so.\", 'start': 1359.593, 'end': 1389.125}, {'text': \" That's the very, very, very human and structural reason why you have 10% of your portfolio in something that doesn't care. What it doesn't, it's, these are computers. They don't have to retract research statements. They don't have to talk to clients. They decide, you know, three a month ago, they decided crude oil was going down this month. They decided it's going up and they can decide it's going down, down, down again in three months. And, and that nimbleness.\", 'start': 1390.255, 'end': 1418.757}, {'text': \" That ability to change your mind to do it efficiently is a great compliment to a portfolio that is slow, slow moving. And it turns out when you look at the returns over time, the times when it does the best are a year like last year or the great financial crisis. So when you look at this thing over a long period of time, it's an annoying investment, 80% of the time, because it does a little bit better than cash.\", 'start': 1419.432, 'end': 1447.63}, {'text': ' in hedge funds, right? So if cash is earning five now, next three years, maybe you do 5%, 6% or something.', 'start': 1448.542, 'end': 1455.275}, {'text': \" But it's not like a CD, right? They drop 10% in March. They come roaring back a bit. They give it up again. So now the problem, the problem with a strategy. 10% in March, I think I know this anecdotally because they follow the hot trades, and the hot trade was shorting rates. And then rates rallied big time when the Silicon Valley Bank collapsed. Of course, right? These guys hate, if you like waves, you hate whipsaws.\", 'start': 1455.899, 'end': 1480.604}, {'text': \" Right, you hate and that's why they hated a lot of the late 2010s was because every time, you know, the markets would really start to get, you know, get their markets are really starting to go down, the Fed would step in, or something would happen. The the way to think about the strategy is that it makes your life better. Because it does things that you're not going to do on your own.\", 'start': 1480.604, 'end': 1499.268}, {'text': \" Now, when I mentioned being annoying 80% of the time, the problem is, is, is for, you know, you're on a golf course with a guy who's your client and he says, Oh, you know, what's that thing in there? That's a 5% allocation. And how come my equities are up 15%? And this thing is down five. It's hard to explain. It's hard to explain. And so, and, but, but this is, I think the point you said, it's like, if you then say, Oh, because there was a sharp reversal in the wheat markets and blah, blah, blah, blah, blah. And so like, and.\", 'start': 1500.044, 'end': 1529.002}, {'text': \" If I'm a client, I'm like, why am I bothering with that? Right. I mean, this is just stressing me out to what's in there. And for the advisor, it's often stressful because how is that comforting? Right. You buy cheap value stocks and they go down. There's a great narrative around it.\", 'start': 1529.187, 'end': 1547.935}, {'text': \" They're getting cheaper. They're cheaper, right? They're cheap. That's why they can be so tough. Four times cash flow. It's everyday low pricing. You buy growth stocks that go through a hit and everyone can say, oh, remember when Apple, I remember when if you bought Amazon at the bottom, you'd be a trillionaire today. And the thing is, every other asset class has this great language around it. It sounds great. Remember leveraged buyouts? I'm sorry, we're private equity now.\", 'start': 1548.003, 'end': 1577.095}, {'text': \" It's private. Maybe you can join our club. It's just, I mean, it's, it's, you know, hard asset based lenders, you know, knuckle dragging, you know, guys who will break your kneecaps, you know, but, but they'll lend your small, sorry, we're private credit now, right? We're sophisticated, you know, grow, I mean, value stocks used to be called cigar butts.\", 'start': 1577.23, 'end': 1599.623}, {'text': \" You know, they're, they're, they're so cheap that you've got the worst management team in the world with the worst business and you're still going to make money on it. And, but now it's value stocks. Value is good. Growth is good. Quality. Yeah. Oh, I don't invest in expensive stocks. I invest in innovation stocks, innovation stocks, right. Everything has a brand and then manage futures. Okay. Like, like what in that sounds remotely appealing. And they said, Oh no, we're going to, we, we have a better term for, we're gonna call it trend following.\", 'start': 1600.028, 'end': 1627.298}, {'text': \" Well, it's actually how they make money. Don't you know a million guys who bought things at the wrong time because the trend was going in their favor? It's funny. So when I first talked about this, do you remember in Trading Places, the movie in the 1980s, there was that scene where everyone's like, buy, buy, buy, buy, buy, sell, sell, sell, sell, sell. And so when I first started talking to people about managed futures, I said, that's sort of an example of the emotional nature of the markets. And I realized I was making people really nervous.\", 'start': 1628.463, 'end': 1655.969}, {'text': \" Because they weren't thinking that they were Eddie Murphy and Dan Aykroyd controlling it. They thought they were the guys in the pits. You know, who are the ones who are getting caught up in it? Because they were thinking about like, you know, yeah, well, I bought that house at the very peak of the market, or I, you know, I bought that stupid stock after it got up 500%. And so trends are not inherently good.\", 'start': 1656.307, 'end': 1677.94}, {'text': \" You want the good trends. So one of my projects, because if you look at our ETF, and again, you talk about hedge funds have not been great investments in general because of the fees and other things. When people have tried to put hedge funds into ETFs, it's been a catastrophe. It's been a disaster. Why? And why is that? Because most things that hedge funds do don't work well in an ETF. If you've got a guy who has some deeply illiquid assets,\", 'start': 1678.666, 'end': 1707.134}, {'text': \" And then you drop them into an ETF and everyone can see those positions every day. Or a guy who owns, you know, who's who owns 20% like it's just hedge funds don't want the transparency, the constraints of running it, you you can't invest in a lot of things like so in general, what people try to do is say, Oh, we'll take this kind of established hedge fund, try to drop it into an ETF. And no matter how low the bar is, for the actual hedge funds, these guys skate way underneath it.\", 'start': 1707.708, 'end': 1733.088}, {'text': \" how low the benchmark is. Yeah. Like, like in others, if hedge funds do five, you know, somehow these guys managed to do two. And so, and, and it's, it's, so think about replication, which, which I think is unique in the liquid alternative space is it's blindingly simple. If these guys make 10 in a hedge fund, you're paying four and the clients are paying four and they're ending up with six. Maybe you can replicate eight.\", 'start': 1733.56, 'end': 1762.417}, {'text': \" or nine with cheap, low cost instruments and end up with seven or eight, as opposed to six. It's, it's, you're not going to get everything that they do, but, but you don't have to when you're that much cheaper. So what we're trying to do is basically how, if you look at our ETF and the managed future space this is an ETF that's, that's delivered 900 basis points of alpha or something relative to the S and P 500.\", 'start': 1763.092, 'end': 1791.644}, {'text': \" since we launched, since it was launched in 2019, that's like millennium Citadel, like that, but it's a daily liquid ETF with an 85 basis point expense ratio. We are right now, if you look at the size of the ETF, we are 0.01% of the ETF world. So you have something that if you package it as a hedge fund should be a $50 billion hedge fund.\", 'start': 1791.712, 'end': 1819.522}, {'text': \" But in the ETF world, it's people haven't even begun to make these kinds of investments. And part of the reason is this messaging. I talked to people who have spent their lives buying ETFs, and they have never a role on a futures contract is an alien concept to them. A futures contract is an alien concept to them.\", 'start': 1820.332, 'end': 1843.99}, {'text': \" that's pretty bullish, actually, that like, I, you probably want to invest in things that are hard to sell, whereas things that are easy to sell, everyone's in them, you know? Well, so so one of the other things about replication, because they're like the we have this expression, like, like, you know, size is the enemy of performance, or asset growth is the enemy of performance in hedge funds. So so like the great hedge fund strategy du jour has been these multi strategy hedge funds, which are have been\", 'start': 1844.092, 'end': 1866.164}, {'text': \" These guys look like magicians. They basically given you what has looked like a 12% CD for the past several years during a period of time when you weren't making any money in your bond portfolio and most hedge funds weren't doing well. Right. You may be underperformed the S&P 500, but if it's such low, such low, so much lower volatility that the Sharpe ratio is very good. Exactly. I mean, it's, I mean, these, these things are, have been unbelievable, but\", 'start': 1867.7, 'end': 1895.982}, {'text': \" You know what, I just wrote this note that we'll put up next week. Kind of some serious questions that people should be asking about it. So one is these guys have all grown massively, right? They've all hired tons and tons of people. And you start asking questions. It's like you were so good when you were one fifth the size. And you had these great team of guys. Is team 56 really as good as team four?\", 'start': 1897.18, 'end': 1923.455}, {'text': \" The crazy thing about these things, and a lot of, you know, I think anybody who wants to be good in the investment business should spend a lot of time looking at history because you have to kind of understand how we got here. So a lot of these, a lot of these firms can actually pass through the cost of their employees and their bonuses and their etc. back to their investors. So these funds, when I say it's a 12% CD, they may be making 20 or 24 and taking half.\", 'start': 1924.872, 'end': 1953.644}, {'text': \" Okay, but now think about how strange that is, right? I'm not bearing the cost of my employees. My investors are. And I see a guy out there who looks like he's great at what he does. And I and the next guy, we're not footing the bill for it. It doesn't come out of our pocket.\", 'start': 1954.589, 'end': 1972.662}, {'text': \" Like, there is no better recipe for overpaying. Yeah, you're going to hire people up until they're marginally useless, because even someone who has a tiny bit of value to add, you're not subtracting their salary because the client's paying for it. And so those are the two big issues. And the third big issue is these are very leveraged funds, right? And so they've done so. But it's sort of an interesting historical note. This is the 25th anniversary of long-term capital.\", 'start': 1974.704, 'end': 2002.514}, {'text': ' And so if you want to, if you want to understand risk management in 2023, read about long-term capital, because everybody who does risk management and anything on the quant side, you know, grew up reading about this as, you know, if you study history, you study world war II, you know, you study the Vietnam war, whatever. So, so, so the interesting thing about it is, is that these, these hedge funds in order to generate those kinds of returns, take a dollar of your capital, turn it into five or 10.', 'start': 2002.632, 'end': 2032.062}, {'text': \" and then manage to make a little bit of money on each of those dollars, take about half for themselves, and then leave the rest for you. And that's like if you looked at like a Goldman Sachs trading business from the 1980s or 1970s, it's the same kind of idea. We're going to have great information, great execution, great borrowing costs, etc. We'll make a 40% return on equity, we'll take half and pay our guys and everyone is happy.\", 'start': 2033.243, 'end': 2057.459}, {'text': \" But the prop trading, a lot of those move from banks to hedge funds. Exactly. Right. And so part of the structural opportunities, these guys are doing things that those guys, you know, used to do in the past. But But here's a question, right? If you're leveraged five times, isn't that sort of a problem today?\", 'start': 2057.594, 'end': 2073.119}, {'text': \" You're borrowing costs, right? We all know if you were a mediocre real estate investor in the 2010s, you are a rich guy who looks like a genius today. Everybody, private equity firms, like zombie companies, like people who were borrowing money in the 2010s on size did really well. The question going back to the point that I was saying is like, if rates, if you think rates are actually going to stay\", 'start': 2074.165, 'end': 2098.212}, {'text': \" where they are. And most hedge funds actually do, right? Most hedge funds think we hit an inflection point, we're in a new world, de-globalization, bop, bop, bop, bop, bop, bop, bop, bop, all these different things. Governments apparently, you know, the restraining governance has gone out the window, whatever. So, but if that's the case, how does it affect hedge funds?\", 'start': 2098.212, 'end': 2117.838}, {'text': \" who have these huge borrowings, shorter borrowings, basically. Yeah, interest rates at 5.5% make leveraged investments much less attractive. And that goes for stocks, it goes for real estate. Interestingly, like if you had a parallel shift in the yield curve, maybe it doesn't for like carry trades or leveraged credit investment or banking, but we don't we have an Inverley yield curve.\", 'start': 2118.952, 'end': 2138.071}, {'text': \" Well, it's like in theory, right? There's like academic theory is like, well, you know, if it would have been a flat yield curve and everything's a 5% and then, well, of course, my return on equities is going to be 5% plus the equity risk premium. So, of course, I'm going to get a 10% return on equity over the next 10 years. Anybody who's hanging their hat on theory working today really needs to find a different job.\", 'start': 2138.071, 'end': 2163.265}, {'text': \" I mean, we had in the 2010s, right? I mean, bonds were earning, I mean, cash was earning close to zero, like let's say 1% over that period of time. And the S&P was up like 13% a year or something. There's not a person in the world who predicted a 10-year period with a 12% equity risk premium, you know, a modeler who kind of predicted the world would look like that. Anyway, so it's like the tiniest turn has gone the other direction. So what I basically read, I said like,\", 'start': 2163.721, 'end': 2189.573}, {'text': \" What's the answer to this? What's the answer you get? Now, the funny thing, maybe not so funny, is that there has been so much demand, people lined up around, institutions lined up around the block to throw money at these guys, that these guys were not only increasing fees, but they got guys to tie up their money. So you tie up your money for five years, and even then, if you want to get out, you can take out maybe 5% a quarter.\", 'start': 2189.945, 'end': 2213.975}, {'text': \" And when we're talking these guys, we're talking about hedge funds that have some kind of vault targeting strategy where they they're going to make that 12 percent CD. They're actually making 20 percent, but they're borrowing at five X. Those are the hedge fund is like a million different things. I mean, they're called multi strapped hedge funds. Yeah. Yeah. And, you know, the FT has covered them recently. A guy, Bloomberg, maybe Sean Kumar covers them a lot. These are the these are the modern day George Soros is Ken Griffin's one guy, you know, Millennium Citadel.\", 'start': 2214.363, 'end': 2243.017}, {'text': \" It just seems to follow the kind of historical playbook that when you've got institutions lining up around the block, and then these guys are, you want to invest with hedge fund managers who are rapacious, ruthless capitalists, right? That is their job description. You want them to be that. But there also can be rapacious and ruthless with their clients.\", 'start': 2243.017, 'end': 2262.743}, {'text': \" And so they've got their clients now to have locked up. Some of them used to be able to get your money out relatively frequently. Now, because of the demand, anybody who stayed is not getting their money back for at least five years. And that's a key problem, Andrew, because let's say the best, you know, a lot of tail risk managers who I like, but the best tail risk manager in the world, let's say hypothetically, where they have almost zero bleed, like how much money they make, how much money they lose in bull markets.\", 'start': 2262.743, 'end': 2288.95}, {'text': \" And then when you have a crash, they make 60% or 100%. How much value is that if you're not allowed to rebalance when the market crashes and your tail risk sleeve of your portfolio goes up, but you have a lock up, right? Isn't the whole point that you're supposed to rebalance on that March 23, 2020, when the S&P is at 2300 and the VIX is at 85.\", 'start': 2288.95, 'end': 2312.339}, {'text': \" But but no one actually does that. And if you're not using an overlay, you literally you're not even allowed to. Or, you know, it's going to be five years until you see that money. Right. Oh, look, I mean, I mean, rebalancing is a nightmare. Right. And I mean, so the thing is, you know, we look back at these periods in history as though we had a clue what was going to happen. I can tell you on March 22nd, you know, good luck getting people out from underneath their desks. Right. It was we were in I mean, the Treasury market\", 'start': 2312.339, 'end': 2341.617}, {'text': \" Was it wasn't trading? I mean, we were on this market was not working. The market was not working, right? I mean, this is this was by ammo and bottled water and head to the mountains, right? Yeah, and toilet paper. And so yeah, and, you know, go to Burning Man. So, you know, so we look back at it, and we want to believe that we would have known what to do. And, and the reality is,\", 'start': 2342.377, 'end': 2362.998}, {'text': ' Like there are the guy that I worked for this guy, incredible investor named Seth Klarman, Warren Buffett. Like there are a handful of guys who wait for the day as Warren Buffett said, when the economic storm clouds darken and rains bars of gold. A funny thing happened in 2020. The economic storm clouds darken, gold bars coming down, and there are 150 people out there.', 'start': 2363.487, 'end': 2390.453}, {'text': \" With big tops, it's people have learned to buy dips. People have learned to step in, to preserve capital. So, you know, the business will keep changing over time. The reason again, going back to the strategy, the reason I like it, right, is that, is you need something that you can put in your portfolio today and just leave it there, right? That's the ideal diversifier. If you are a financial advisor,\", 'start': 2391.55, 'end': 2420.947}, {'text': \" is something you can put in your portfolio today, talk about it when you want to, and never talk about it the rest of the time. Because you will go through periods where it may not be fair, but you'll go through a period where it's not doing as well as something else in the portfolio and people focus on it. But that's really, but that's, but, you know, I know a lot of really, really smart people who gave up on this space in 2020\", 'start': 2421.014, 'end': 2446.748}, {'text': \" just before a very good 2021 and a decade of outperformance, 40 points of outperformance, 50 points of outperformance in 2022. Didn't they do well in 2022 because they were so short oil or no?\", 'start': 2447.17, 'end': 2462.712}, {'text': \" Look, this is this is sort of interesting about replication is that is that we have this thing. It's like, get the big trades right. Right. And so they were betting that interest rates would go up through by shorting treasuries. They were long oil in the first quarter, sort of luckily, honestly, that when the Ukraine war hit, they ended up making a lot of profits. Then they kind of struggled with oil the rest of the year. And\", 'start': 2462.712, 'end': 2487.738}, {'text': \" In a world of rising rates, this is slightly technical, but generally, if you're going to raise rates, but another country is not going to raise rates, your currency is going to go up. And so as the Fed started to raise rates, there's this just very bizarre thing where Japan, because of their situation, is like, you know, has a white knuckle grip. They are never going to raise rates as far as they can. So there was a huge move in the end versus the dollar.\", 'start': 2488.632, 'end': 2512.561}, {'text': \" And in fact, yen, I think, you know, we think was by far the most profitable individual trade. But you know, when people talk about this space, they talk about, you know, well, did you see how much you know, natural gas went up? It doesn't matter. I mean, doesn't it's, it's, it's, it's like saying, it's like saying, look, if value stocks are going to outperform growth stocks over the next 10 years,\", 'start': 2512.561, 'end': 2529.031}, {'text': \" You don't need to do a lot of brain damage trying to find the one value stock that's going to do better. You can get, get value, right. Versus growth or growth, right. Right. Versus value. And that's all, that's always been a truism of hedge funds. And that's why replication, you know, for guys who've been in the industry for a long time is very obvious to people. Because if you were an equity long short manager in 2000 and you were long, small cap value stocks and short large cap growth stocks against it.\", 'start': 2529.875, 'end': 2554.833}, {'text': \" your longs went down 40%, right? If you're a long only manager, you were not looking good, but your shorts went down 80. And so when the world went to hell as an equity long short guy, you preserved capital, right? It was a magical\", 'start': 2555.93, 'end': 2573.092}, {'text': \" outcome, which really sets the hedge fund industry off in the next 10 years to a necessary thing in everybody's portfolios. Because anybody staying in 2005 looks at this and says, oh my God, if they did it once, of course, they'll do it again. Now, of course, they didn't do it again, right? 2008 came around, they should have gone down 14 based upon their level of risk. Equities went down about 40, right? So they should have gone down about 14, then went down 20, 22.\", 'start': 2574.188, 'end': 2600.125}, {'text': \" It was an embarrassment. As the hedge fund industry has matured, it's not giving your money to that guy that you met who is super smart, who you just trust to be this great steward of your company. Look, the guy that I worked for was, if somebody said to me, here's a billion dollars, give it to one guy and you can't look at it for the next 30 years.\", 'start': 2601.307, 'end': 2625.826}, {'text': \" You give it to him, right? It's easy. He'll figure it out. Because you know what? If I have a billion with him, he's got three. And there's no human being on earth who likes to lose money less than that guy. But how would you know who's going to be the next great guy? How would you know? Because that guy, even that guy has gone through periods of massive underperformance because he's in areas that weren't hot at the time. He wasn't big. So look, investing is really hard.\", 'start': 2626.535, 'end': 2654.041}, {'text': \" Picking managers and the guys who are going to do well is a largely hopeless task on a systematic basis. And so picking the asset classes with any level of confidence that are going to outperform what I'm going for, this is the year of the bond.\", 'start': 2654.412, 'end': 2670.46}, {'text': \" In January, it was a no-brainer that you should be buying bonds, because not only because the Fed's going to taper, rates are going to come right back down, and that's going to be the huge tailwind versus the headwind. There's going to be a recession. Bonds are going to outperform stocks. Stocks are dead money. And also, if you are in stocks, or if you're overweight stocks, make sure that you're not in large cap growth stocks like Apple or Forex.\", 'start': 2671.388, 'end': 2695.317}, {'text': \" Microsoft or Nvidia, you want to own small cap value stocks. Yes, of course. Yeah. Hasn't this been hell for hedge fund managers who are, I guess, I suppose, correct me if I'm wrong, kind of always in aggregate going to be underweight Apple and Nvidia, right? They weren't. They weren't in the 2010s. And then they bought it. Actually, there were some guys. There's a group of hedge funds called the Tiger Cups. And I knew some of the guys. And\", 'start': 2695.317, 'end': 2722.57}, {'text': \" In 2014, I was sitting with one of the guys, and I was looking at his portfolio, and his biggest position was Apple. And I was like, seriously? How do you have an edge in Apple? When you and I left business school years ago... Why is someone paying you $2.20 to own Apple? Don't Apple. And his argument, and this became... His argument was basically, when Wall Street analysts look at Apple,\", 'start': 2722.57, 'end': 2750.684}, {'text': \" They, they don't understand compounding. They don't understand that Apple, even though it's big, can grow 30% a year or whatever the number is. So there'll always, there'll always be a 0.3 years out where they start to get it wrong. And, and we do these things, we go out much older, we're smarter, but it was basically claiming that they had this edge with it. And they were right. I was super skeptical at the time, but they were right because what they recognized was that those big tech companies were a different breed of animal.\", 'start': 2751.595, 'end': 2780.587}, {'text': \" They had a, you know, I mean, Warren Buffett had having missed investing in Google. He has this great quote about Google. It's like somebody invented a cash register, put it in San Francisco. And every time anybody goes online anywhere in the world, the cash register rings. Anyway, so back to my point about hedge funds is they actually did catch that early. Some of the guys caught it early and they had, and the, you know, the, the thing was, it didn't matter whether they own Microsoft, whether they owned whatever, whatever, whatever they got, they got the whole category, right? These were a different breed this year.\", 'start': 2781.565, 'end': 2811.417}, {'text': ' No, no, no. This is in the 2010s. This is like 2014, 2015. Hedge funds, which had shunned big growth stocks since the early 2010s, started to buy into the idea that these things were these cash flow generation machines, the world of which we had... Compounders. Yeah. But then once the crisis hit,', 'start': 2811.535, 'end': 2833.591}, {'text': \" And the stocks went up even more. A lot of them said, well, OK, not so funny anymore when these five stocks are now 40% of my portfolio. I'm going to dial them back. And they didn't want to buy Peloton because Peloton was ridiculously expensive and a weak business. Well, whatever it was. And they looked around and they said, oh, Carnival Cruise Lines is down 98%. And this theater museum. So there was actually a big shift at the margin to value.\", 'start': 2834.587, 'end': 2863.578}, {'text': \" And so think about replication is like when you're looking at these guys, and we do replicate equity long short as a strategy, we have a separate ETF that does that, you know, but what you care about is, are these guys looking at stocks in every day, deciding that those tech stocks are have just gotten too expensive. It's not worth it anymore. Have they have they made a move from, I mean, you know, the S&P 500 went up 250%. In the 2010s, NASDAQ went up 425%.\", 'start': 2863.747, 'end': 2889.768}, {'text': \" Europe went up 80% cumulatively, and emerging markets went up 50%. That throws, you talk about theory versus reality. You imagine those like security market lines. Oh, you take more risk, you'll make more money. Uh-uh. And it's not supposed to go on for 10 years like that. So hedge funds started to say, you know, I bet the 2020s,\", 'start': 2890.105, 'end': 2912.836}, {'text': \" the valuation convergence is going to happen in a higher interest rate world. The fact that Europe has more, I mean, Europe has underperformed for decades. It has been the cigar butt of asset allocation models. Right. And but it's going to change. It's finally going to change. They're going to come back. So earlier this year, kind of a big trade among hedge funds was this is the revival of Europe. And then they got run over. So it's it's a really, really humbling environment.\", 'start': 2914.085, 'end': 2942.384}, {'text': \" For a lot of people, you can get it right one month, but getting it right all year long. I just, I don't know of anybody who got last year, right. And then turned around and flipped and got this year, right. And then, uh, even like the, those, those guys who made a fortune on tech stocks, um, you know, they go down 50 and they're up 20 this year. That's not enough. You're still down 40 at two and 20.\", 'start': 2942.671, 'end': 2968.034}, {'text': \" Yeah, you're talking about people who they invest in more than Apple. They take, you know, they're investing in unprofitable companies as well, not just Microsoft Apple. There was a band of hedge funds that I was describing, these Tiger Cubs. Yeah, yeah. Who had huge exposure to, you know, to these large U.S. large companies. The noons of the world, let's put it that way. The Googles of the world, the Google, Microsoft, etc. And they actually pulled money from\", 'start': 2968.844, 'end': 2993.026}, {'text': \" So they didn't capture the rebound. Basically, they get pummeled last year. Historically bad numbers last year. Then they're like, okay. And they even some of the smartest guys in the business, actually, some of them actually bought into the idea that this is going to be dead money for a long time. Yeah.\", 'start': 2993.195, 'end': 3013.343}, {'text': ' And then, and then, and then look, I mean, maybe one or two of them happen to have Nvidia in their portfolio, but the whole chat GPT thing was, was really came out of left field.', 'start': 3014.288, 'end': 3023.452}, {'text': \" So when it comes to rebalancing, how often does the studies, the research suggests that is optimal for trend following or managed futures? If you have, let's, I mean, you know, 60% bonds, 60% stocks, 30% bonds, 10% managed futures. You have a year like last year where stocks and bonds both had a horrible year, but managed futures did really well.\", 'start': 3024.903, 'end': 3047.498}, {'text': \" Presumably, you take some money out of managed futures, put them into stocks and bonds. How often do people do that quarterly? What does the best practice for that? Do it in the simplest and most straightforward way. Don't get cute. The problem with managed futures is like you say with tail risk. People want tail risk on March 18, 2020.\", 'start': 3047.498, 'end': 3066.938}, {'text': \" when it's already up a bajillion percent, right? They didn't want it three months before. And so, so I just think you have to be honest with yourself. Am I likely to get the timing right on it? And, you know, the problem is like, like gambling, like day trading, like all these different things. It's fun.\", 'start': 3067.377, 'end': 3086.564}, {'text': \" Right. And so you should have fun with it's your money. You should have fun with it if you want, but don't fool yourself that you're going to know how managed futures is going to do in the next month, even how equity is going to do in the next month. And so for the core of your portfolio.\", 'start': 3087.138, 'end': 3104.131}, {'text': \" The ethos for the next 10 years should be, I'm going to have a meaningful slug in equities. I'm going to have a meaningful slug in bonds, however you describe it. And I'm going to have a meaningful, but not as large slug in managed futures. And I'm going to basically, that's my bet for the next 10 years, because each one of those things.\", 'start': 3104.823, 'end': 3124.178}, {'text': \" is not static, right? They talk about an equity allocation is static. The S&P 500 is not static. And so the S&P 500, if you, you know, we started talking about this idea of beta, we don't know how it's going to do. My guess is it's going to do a lot less well over the next 10 years. Just statistically, you can't keep compounding at 13%. Sorry, S&P. S&P 500. It tends to be very tech happy. It probably sounds bad. But, but, you know, those are,\", 'start': 3124.465, 'end': 3153.203}, {'text': \" Being honest with myself, I'm not likely to be more than 53% right, that I'm going to get that right. So don't bet the house on it. You have companies out there, whether in the Europe, US, or whatever, where you've got guys running them who are trying to build and increase value. Everybody wants that to happen. You've got bond investors out there who are trying not to lose your money and find the things that earn more money over time.\", 'start': 3154.638, 'end': 3181.57}, {'text': \" And in managed futures, you've got guys who are trying to identify the opportunities that you're not going to see in buckets one and two. Right. But Andrew, so with managed futures, you do have a positive carry, a return greater than zero, which is much more than tail risk, people can say. But you also have that\", 'start': 3181.638, 'end': 3200.892}, {'text': \" zero correlation, or in some cases, when other assets go down, they go up. Isn't the whole benefit of that that you can take that cash is more valuable after assets go down because you can buy them back at lower values? That's the tail risk pitch. That's why they say, oh, you\", 'start': 3201.01, 'end': 3216.215}, {'text': \" So isn't part of the benefit of managed futures that you take money out of managed futures the year after it did so well to buy stocks and bonds at lower prices? I'm all in favor of rebalancing over time. But all I'm saying is, I would just say, don't get cute. No one is going to pick the top of assets and markets, et cetera. And don't even think that you're going to.\", 'start': 3216.957, 'end': 3243.755}, {'text': \" because that's just going to lead to a lot of brain damage and anxiety and disappointment. If you're going to do rebalancing, which you should do, do it quarterly, you know, and maybe change things at the margin, but take the market timing decision away from the view of what you want that portfolio to look like and how you want it to grow and how much risk you want to take over the next 10 years. Because the rebalancing, if you get cued on the rebalancing side, you can blow up the rest of the equation.\", 'start': 3244.058, 'end': 3273.59}, {'text': \" I mean, it's sort of interesting. So one of the things that I learned from this guy, Seth Klarman, years and years and years ago is don't do stuff just to keep busy. And so the year before I joined, it was sort of funny in that they... And what year did you join? 1994. So a much, much, much smaller firm, just a little boutique firm. I got recruited by some of my professors at Harvard Business School who had been actually founders of the business back in the early 1980s.\", 'start': 3274.805, 'end': 3303.695}, {'text': \" And and I didn't know anything. I didn't know what a hedge fund was. I didn't know who the guy was or whatever. But I went in and they basically beat the crap out of me for two hours in an interview, asked me all sorts of hard questions. And I thought it was kind of fun. And so I ended up going to work for him. But but, you know, but the thing was, like, if you're invested in cash, right, if you're it's it's a high hurdle to take money out of them.\", 'start': 3303.695, 'end': 3325.413}, {'text': \" That's OK. The problem we had in the 2010s was that there was this concerted effort to make sure you didn't hold that safe asset. Because it was yielding nothing. Because it was yielding nothing. And so that's what I mean by artificially low. A risk asset shouldn't earn zero over time. But what Seth would say is don't buy that stock just because you want to go buy something and you want to feel busy or whatever. You've got to be able to justify it as to what you're not doing.\", 'start': 3326.459, 'end': 3354.421}, {'text': \" And, and so the way that we've approached what we're sort of known for in the, um, you know, in this world of where people try to take hedge fund strategies to put them into mutual funds and ETFs is we've, we've only made a few bets over time and they've all worked. And by that, I mean, you're talking about that post. No, no, I'm sorry. I'm talking about like our, our, our firm, our firm today is that, is that we've, we've, we've identified a handful of strategies where you can do better than actual hedge funds.\", 'start': 3355.18, 'end': 3384.678}, {'text': \" But in an ETF or in Europe, we do these things in mutual funds. But with daily liquidity, low fees and all sorts of like, like really salient features. And and but we haven't done our stupid things. We didn't go create products that that that ended up not working five years later. And we had an egg on our face. And you know, the asset you didn't create products that have no real use case, you're creating a product to create a product. Yeah, look, I mean, to me that to me, the answer is question, can I do better than hedge funds?\", 'start': 3385.252, 'end': 3412.42}, {'text': \" and offer you daily liquidity, let you invest $100 as opposed to $5 million in it? Can it be, can I not, you not have a K-1? Can you have all of these things that make your life at the margin better and help you with your broader portfolio? There's a huge bias in this business to just come up with new things. What's the new thing? What's the new hot idea? Most of our terrible ideas and you waste so much time\", 'start': 3413.129, 'end': 3441.547}, {'text': \" like, oh, we launched a new product. Well, I mean, you look at the ETF world, it's like everyone's launched a new product, this product, this product, this product, three years later, five years later, half of them, three quarters of them will be gone. And and what it does for the for the allocator, for the advisor, is you're bombarded with these ideas.\", 'start': 3442.897, 'end': 3459.721}, {'text': \" you're bombarded with, you know, somebody coming in and saying, oh, you know, we just launched the, you know, tail risk fund where we're going to earn 10% a year. And then also happen to make 50% a year at the right time. And then if here's some hypothetical numbers that we put together and, and that's, you're, you're really much better off just focusing your time and energy on a limited number of things that make sense. And, and, and that's, that's why we're a bit different. I run a quantitatively based firm.\", 'start': 3460.328, 'end': 3489.505}, {'text': \" But we have avoided a lot of the things that people have done in the quant space that sounded great at the time, but then ended up looking bad. Like, what's an example of something that sounded great in 2013 that just didn't? Oh, my God. So 2013, 2014, these quants basically came out with things and said, we're going to do what hedge funds do. We're going to do it cheaply. But they did in this incredibly mind numbingly complicated way. It was an area called alternative risk premia.\", 'start': 3489.758, 'end': 3518.176}, {'text': \" And the best and the brightest jumped on this. Every invested firm jumped on it. People launched mutual funds. They did all sorts of other stuff. And we looked at it. It sounded great. Back then, you weren't earning much money in your cash. It was supposed to be a 6% return, not a lot of volatility, no correlation to anything.\", 'start': 3518.699, 'end': 3539.472}, {'text': \" And it'll be daily liquid, daily liquid, not too expensive, other stuff. And, you know, can you explain for our audience who, you know, are human beings and not computers, like what what is the rough if you know, if managed futures is long, things are going up, short things are going down. What is alternative risk premium? So these are these were these were quantitatively based strategies where you were making these\", 'start': 3539.759, 'end': 3559.823}, {'text': \" long-term bets based upon what had happened over the past 70 years. So, oh, our model says over the past 70 years, value went up. So we're going to buy value in short growth. Disaster. Disaster, right? We're going to do this particular currency pair because it's worked better over the past 20 years.\", 'start': 3561.224, 'end': 3578.892}, {'text': \" and disaster disaster. So and so. So I was looking at it. And because I was saying, God, I mean, if we could, if it worked, right, think how great it would be, you could get these, you know, let's say other sources of alpha, whatever. And I kept raising my hand that I was saying, like, like, you ever see the movie big? Yeah, yeah, yeah. So I remember there's a scene where Tom Hanks is they're talking about this new product or whatever, whatever. And he kind of reasons like, I don't get it.\", 'start': 3578.892, 'end': 3605.386}, {'text': \" And they're like, what don't you get it? I was that guy. I was like, I don't get it. Why didn't you do this 20 years ago? If it's such a great idea, who actually started doing it 20 years ago, past performance is not a predictor of future returns. You know, I mean, I put up things on Twitter all the time. And you know, oh, the last time the unemployment rate up when this time on a historical basis, this is what happened 12 months later. I'm just some guy on Twitter. I'm not putting people's money. You know what I mean? Like, the standard is high enough for me to it's an interesting thought. But\", 'start': 3606.162, 'end': 3634.613}, {'text': ' You know, things change, regimes shift change. Before 1997 or 1993, you tell me, I mean, stocks and bonds traded in line with each other. It was only a kind of a recent phenomenon that stocks and bonds, you know, bonds were a hedge for stocks. Things change all the time. So look, a lot of academic finance. Rising rates are good for banks, right? Nope. You know, sorry. No, no, no. So a lot of academic finance is insane. Yeah. You have guys, first of all, think about the incentive structure.', 'start': 3635.66, 'end': 3663.537}, {'text': \" you Jack, you just got your PhD, or you're getting your PhD at University of Chicago, it's time to write your thesis. And you work on it for six months, you don't find anything interesting. That's not your thesis. You need to turn the dials some more. Okay. You weren't looking at sunspots, you need to incorporate sunspots. Oh, you know what? And so what happens is the whole academic finance business is this truly insane game of backtesting.\", 'start': 3663.605, 'end': 3692.427}, {'text': \" Right. So I wrote on like the value factor, right? So, so I, like, I worked for one of the great value investors. The value factor, as people thought about the time was debt at the time. Those, the, this mythical company that was trading its stock market capitalization was less than its cash at the time didn't exist.\", 'start': 3692.883, 'end': 3713.352}, {'text': \" Not because- 1994. 1994. Yeah. This is two years, by the way, after the Nobel winning prize paper from Eugene Fama and Ken French came out basically talking about the value factor. And what they did was amazing. But what they described was the past. And by the time they published it, the world had already changed. I knew this because I was kind of nerdy. I almost went into the PhD program at Harvard Business School because I thought, you know, God, it'd be really fun to study all this stuff. Look at all these like, you know, you can study seven years of the markets and find all these interesting truisms.\", 'start': 3713.74, 'end': 3743.457}, {'text': \" And what I realized very, very quickly was that it wasn't that value investors wouldn't have bought it. They hadn't changed. It's that the stock didn't exist anymore because LBOs, the same guy who was looking at a stock and saying, wait a second, if this is trading less than its cash, why don't we just borrow money, buy the company, take the cash, give it out. We'll have a company for free.\", 'start': 3744.115, 'end': 3769.512}, {'text': \" Strategic acquirers did the same thing. The whole M&A business. So M&As, leveraged buyouts, all this stuff happened. So all the companies were gone. So the Warren Buffett stocks that he got rich in the 1950s with his clients, those no longer existed as value stocks. Companies that you could buy that literally had more cash in the bank than the market cap. So now instead, those value stocks no longer existed. So instead, a value stock was just something that traded with the price-to-earnings ratio of five. But there's a reason it trades like price-to-earnings of five.\", 'start': 3770.103, 'end': 3797.693}, {'text': \" It had a lot of problems. Right. So what happened is actually there was this pivotal shift in the value investing world in that they gave up on the buy. Well, let's buy cigar butts with a margin of safety because they didn't exist. So what they started to become, they would say, like, we're going to buy good companies, but at reasonable prices. But the way that they would calculate that was to say, what if KKR bought it?\", 'start': 3797.795, 'end': 3825.301}, {'text': \" What if Carlisle bought it? What's the analysis they would do? Well, the analysis they do is you model it out. We have computers now. You can build an Excel spreadsheet, and you can say if the cash flow goes like this, and you discount it back and come up with the cost of capital, you can decide, ah, the stock should be worth 20 today, and it's trading at 14. Therefore, it's cheap.\", 'start': 3825.993, 'end': 3846.597}, {'text': \" But when you do that, right, when you're looking at the company, at the balance sheet of the company, these terrible company, but with lots of good assets on it, you don't care about management. You know, they're bad. Because if you're talking to them, they're terrible at their jobs.\", 'start': 3847.34, 'end': 3863.759}, {'text': ' the the here you care, because these guys have to make wise decisions and grow and what you know, do they take the money they buy stock, should they go buy another, you know, try to build another stupid factory. And so, so the value business changed in the 1990s, from an asset based business to a company space business. And, and anyway, that that had sort of profound implications. But when I looked at the space, the alternative risk premium space,', 'start': 3864.282, 'end': 3890.236}, {'text': \" I was talking to a bunch of guys who are much better quantitatively than I am with terrible unrealistic business judgment. Because they were basically saying, I remember talking to one guy and he said, well, the underlying strategy itself is not really interesting, but we found a way it only goes up when the S&P, we found a way basically, we know when we came up with a signal, when the S&P is going up or down, and we know it's going to go up or down over the next week, that's when we're going to invest in the strategy.\", 'start': 3891.097, 'end': 3919.227}, {'text': \" And I'm like, if you if you if you think you know, when the S&P is going to go up or down in the next week, why are you bothering with this? My friend, you have a much better career opportunity, just going out and, and managing some money. So but look, let me that's, it's what makes this business fun. It changes a lot over time. You know, you're dealing with\", 'start': 3919.953, 'end': 3941.452}, {'text': \" human beings who are we're all trying to make good decisions in what we do. But but, you know, there are lessons you can learn to to to avoid making. The big mistakes and a lot of the asset allocation business, a lot of the things is is people making every now and then you could look at things and say, like buying a 10 year. Double A double B rated bond at a one and a half percent 10 year yield when the company was\", 'start': 3941.603, 'end': 3970.24}, {'text': \" not so great back in 2021, unless you were forced to buy it because you had some bucket that said, I must own this certain amount of AA bonds, that's one you should walk past. Or the 10-year treasury. I mean, the real overvalue thing was the risk-free rate, not necessarily the credit spread. Maybe you're right. Yeah. Credit spreads were tight. But the thing is, though, if you were saying that in 2021,\", 'start': 3970.308, 'end': 3998.557}, {'text': \" And you were saying it in, you're probably saying in 2019, and you were probably saying in 2017. Yeah. The people who were at events who were like, oh, Andrew, you've been saying this for years. So you may be right. But look, I mean, sometimes, I mean, it's, I think it's like, just be humble. I mean, about this and spread your bets and, and think through, one guy described us, he was talking about kind of our approach to hedge funds.\", 'start': 3998.827, 'end': 4025.877}, {'text': \" And, and he says, he said, I don't know whether to say you guys, it's like, it's either common sense, or the thinking man's approach. Because you're not saying you have the answers to everything. But but you can, you're pretty good at figuring out when the odds are stacked against you, in a very, very meaningful way, and not doing it.\", 'start': 4025.928, 'end': 4050.11}, {'text': \" Because we're not working at a bank where our job is to crank out products. We're not working at a quant firm where we have to like, you know, come up with the next new product this year. We heard somebody else is doing a product and we're going to copy and launch, launch it. It's, it's, you know, we want to be known for the guys where we're going to tell you exactly what we're going to do and what we're going to try to do. And we're going to explain to you why we think it's a good idea. And we hope you're our, you're our client today. And in 10 years, we're doing, we're doing exactly the same thing in that fund.\", 'start': 4050.734, 'end': 4078.966}, {'text': \" And, and, and you're happy with the outcome. And that's, that's not, that's a very rare experience in for, for, for allocators because they spend so much time jumping on the thing that just looked great.\", 'start': 4079.692, 'end': 4093.208}, {'text': \" Then it's hugely disappointing. And now they've got to decide, do I sell it? Do I get out? How do I explain to my clients? And all of those dynamics, they're just inefficient and costly. If you're an advisor, you should be spending your time. You want to be able to make an allocation and for it just to work.\", 'start': 4094.508, 'end': 4117.61}, {'text': \" And, and, and that often means simpler, it means more straightforward, it means it means sort of a common sense approach. So we're very strange hybrid in that we are, we have, you know, amazing modeling and mathematical expertise, but we know when not to believe it.\", 'start': 4119.061, 'end': 4136.24}, {'text': \" Yes, you have a really good process, but you're very unambitious in what you try to do because you've seen a lot of people, you know, try and lift that 2000 pound weight and it not go so well. Yeah. And luckily, like you think, I mean, there are really, really smart guys in quantitative finance who have really mediocre and sometimes horrible performance over time. But if you think about how would you get comfortable with a quantitative strategy?\", 'start': 4137.016, 'end': 4166.497}, {'text': \" The longer you look back, the more comfortable you're going to be. Do we really think that the lessons of the 1950s are really relevant today or that the dispersions or whatever,\", 'start': 4167.222, 'end': 4185.464}, {'text': \" the mispricings can be corrected on a time horizon that is sufficient. In other words, yeah, okay, value is going to work by 2030. I mean, if you invest in value from 2030 to 2050, a thousand basis points a year about performance, but it's like, can you wait that long? You know, but, but it's, so this is actually because it's sort of interesting point because, because, you know, you say value.\", 'start': 4186.595, 'end': 4207.199}, {'text': ' Yeah. You cannot find two quants who agree how to define it. Price to book, price to earning, EBITDA. But think about the bizarreness of that, right? We are describing a permanent truth of the market, but we have to change our definition of the permanent truth of the market every two years. That should scare the daylights out of you.', 'start': 4207.368, 'end': 4229.795}, {'text': \" it's in. And so so I think like, like some of the questions that you you, you know, like, if you're if you're an advisor, and somebody is coming in and showing you some hot product, the first question you should ask is, that's great. That's really interesting. Show me every product you've done in the space. And what you thought was your best product three years ago, when you plug in the best product five years ago. It's a tough test. Well, it's because because because the business is built to sell hot dots. Yep. Right, that you're gonna have a firm that's got six products.\", 'start': 4231.634, 'end': 4260.828}, {'text': \" And you're going to only hear about the one that's going up. And often six of the products, three of them are positioned one side of the boat, the other side of the boat. So mathematically, there's going to be one of them that looks like everyone's a genius. Right.\", 'start': 4261.064, 'end': 4275.475}, {'text': \" But like, I mean, I've written about like the asset management industry is designed to fool you into buying things with unsustainable performance. They're called hot dots, right? Hot dots, hot dots, hot dots, hot dots. And Morningstar does great analysis where they basically do the analysis you've described where where the vast majority of money always comes in at the wrong time.\", 'start': 4276.252, 'end': 4296.434}, {'text': \" That's not the asset manager's fault, by the way. In general, it's not the portfolio manager's fault. It's the structure of the industry. It's the- Yes, and I think it's human nature. I just noticed people, if there's someone who's a bear, they're known as a bear, the demand for me to interview that person is at its highest when the market is at a bottom.\", 'start': 4296.434, 'end': 4316.718}, {'text': \" And they hate that person when the market's at a top, you know, or you've had a big rally like right now. And if I were to interview them now, they'd say, Jack, why would you have this person on? When in October, they would love for me to have them. So I think it's, you know, it's everyone's fault. That's what I'm saying. It's everyone and no one's fault. The thing is structurally the industry, like if you want to run an asset management firm, and you go with one fund, okay, you're not hiring a lot of people to do that.\", 'start': 4316.718, 'end': 4344.224}, {'text': \" Cause if you have a rough six months, you have a lot of people sitting around, you know, playing cards expensive. So, so it's rational for you to do Farley one through six. And, and, and then you can hire more people because you always have a product to feed them. Right. That's good from a business perspective for asset managers. And, but just be aware of it.\", 'start': 4345.068, 'end': 4372.81}, {'text': \" Right. So, so somebody comes in and tells you, this is a great product. Say, I want to see everything you're doing. And there are people that we compete with who have one really good product. And, and it just, it's, it's an awkward question. You know, and a lot of the, you know, the human dynamics, this industry is.\", 'start': 4373.519, 'end': 4392.233}, {'text': \" One I know a guy who's one of the kind of like leading luminary hedge funds hedge fund managers And I've known him, you know for over two decades almost three decades at this point. He's really unpleasant to talk to because he The niceties of just letting things go and\", 'start': 4392.588, 'end': 4415.555}, {'text': \" is not wired into him. He is on the front. He is on the balls of his feet. So um, you know, you know, yeah, so anyways, I've rented this place and we're going up there. Why'd you choose that? Did you compare that to something else? What is it? Do you think about it? Like it's you're just like, Whoa, you're always comparing something. Yeah, it's always is it better? Why is that better than this? What is it? But that's what advisors need to do. When people come in and sell them, they're coming to sell you something because they are trying to make money off it.\", 'start': 4416.145, 'end': 4445.575}, {'text': ' A quant comes and writes a paper and sends you their paper about how they found some new magical trade. They are not doing this for the sake of humanity. They are probably consulting for an asset manager. They work at an asset manager. This is not altruism. This is', 'start': 4445.795, 'end': 4463.193}, {'text': \" you know, this is hardcore capitalism. And, and so and that's great in a lot of ways, because they are motivated to find new opportunities to to find new strategies, find other ways to make money, like we want people to be doing that. But you as an allocator, you know, as an advisor, as an investor, just should be prepared to ask some awkward questions. Yep.\", 'start': 4464.053, 'end': 4489.906}, {'text': \" And those questions are like, if this sounds great, why didn't you do it 10 years ago? What did you do 10 years ago? And this is what's strange about replication, is that replication is a very strange strategy, and that we hope we're doing exactly the same thing in 10 years that we're doing today. We want to find a way to do it well. It's not the hedge funds themselves will do different things, but the way that we identify what we're doing and copying cheaply,\", 'start': 4490.345, 'end': 4519.218}, {'text': \" is, is, is, you know, the best outcome for us is, is we found the solution that works. And, and because then 10 years from now, you can say, what were you doing 10 years from now? And he said, we're doing exactly the same thing. But think how strange it is when a quant comes to you and says, we've introduced all these new changes in what we're doing. We found a new way to invest in it. Well, they only do that when it's not working. So, so, okay. So this is your triage.\", 'start': 4519.943, 'end': 4547.703}, {'text': \" right? And the response should not be wonderful. You know, you, you bandaged the, you know, you bandaged, you've lost your, your right leg from the knee down and you bandage it. The question should be,\", 'start': 4548.108, 'end': 4562.502}, {'text': \" How did you lose the... Yeah. Or you're making it a change so that your back tests, which look amazing a year ago, still looks amazing. Right. And so, so, so look, I mean, this, but it's, you know, it's an exciting dynamic business, but I just think it's, I think you can avoid, you can avoid a lot of issues when you just come into these conversations with the skepticism and are willing to ask questions that are a little bit outside of what people normally do.\", 'start': 4563.025, 'end': 4588.135}, {'text': \" I believe in trend following, backtest. If you have a backtest and it's working, keep doing it. If you have a backtest and it stops working, just stop doing it. But the people who have lost a lot of money over the ages, I think they just have one narrative of, oh, this has worked from 1950 to 2019, so it's going to work now.\", 'start': 4588.203, 'end': 4607.913}, {'text': \" And but also think about the dynamics. Right. So I wrote a lot about Cathie Wood when she was when she was going through. And I'm much more sympathetic to Cathie Woods than than than I think a lot of people like you had George Noble on on your podcast once. I mean, the people people are very visceral and crazy to me. The thing about about about Ark and Cathie Wood was that she can't change her mind.\", 'start': 4607.913, 'end': 4634.407}, {'text': \" Right. Kathy, do you think inflation is going to come back? No, it's going to be disrupted away. Right. Yeah. And you you're never going to Kathy Wood. She's never going to sell the technology stocks and buy coal stocks. Number one, that's not who she is. Number two, her clients, that's not what she's hired to do. Right. And so she would not be doing her job. Exactly. Right. So so so don't ask her opinion on inflation.\", 'start': 4635.503, 'end': 4660.462}, {'text': \" Because inflation is unequivocally bad for her portfolio. This is why this is why I so rarely, you know, interview people who obviously you want to talk to people who have skin in the game. But if I interview someone who they are the CEO of an oil company or a copper mine, and I ask them, what do they think about the price of copper? It's like, yeah, I think the price of copper is going to go up. There's a shortage. Surprise, surprise. You know, if you interview someone whose their whole life is about, you know, disruption of tech stocks, they're not going to they're not worried about the Fed raising rates.\", 'start': 4660.462, 'end': 4689.2}, {'text': \" No, and it's not their fault. So but it's like if you ask, you know, you go to Fox News, you know, who's going to win the election. You go to CNN, you know, is going to win the election. Right. It's not it's and it's really interesting about about Cathie Wooden Arc, which was that when she outperformed massively from for a number of years,\", 'start': 4689.487, 'end': 4708.808}, {'text': \" And, and they, you know, they described her in magazine articles talked to her about as this, you know, so the new goddess of investing and the modern day sort of George Soros, etc. And then when she went down, there was a really interesting heuristic flip. Yep. In that, in that, oh, no, no, all disruptive tech stocks are going down. She's representative of disruptive tech stocks going down. So it's not her fault.\", 'start': 4708.91, 'end': 4733.767}, {'text': \" Right. But, and in a sense, she became the theme. Yes. Right. And, and, and that's a really, really, really bad way of looking at it. If you're trying to think about, is this person going to, if you want the theme, of course you should buy Cathie Wood or should buy somebody who's doing, doing something similar. But, um, it's the, but.\", 'start': 4734.965, 'end': 4757.982}, {'text': \" But you're making a bet on the theme. Looking at her as a theme is the right way to look at it. But the characterization of this alpha generator, because the theme is doing well, is like you don't compare the NASDAQ in the 2010s to emerging market stocks and say you're a genius alpha generator. But what it does is it gives people, you know, people need a reason not to sell.\", 'start': 4759.568, 'end': 4788.003}, {'text': \" And when people have made a commitment to a fund, it's their problem now if it doesn't do well. And so ARK had all these firms that had thrown their weight behind her as the genius. And people in the media, not myself, but yeah. And when it starts to go badly, there are haters out there. But the people, she didn't have redemptions.\", 'start': 4789.353, 'end': 4812.809}, {'text': \" Yes. How could you not have redemptions when you have walked into the propeller and stood there smiling, you know, and and and she didn't. And it's because the people who are invested with her had bought into it. Right. And so it takes years of underperformance for people have to have to then be able to say, OK, we're going to switch to something else. We're going to do it. So a lot of the stickiness of this business. And again, we started this whole conversation about, like,\", 'start': 4814.193, 'end': 4842.222}, {'text': ' you know, about what makes hedge funds different is that by and large, the average hedge fund is much better at changing its mind at the right time than because they can. And most of the time and most hedge funds, clients expect them to do that.', 'start': 4842.475, 'end': 4857.427}, {'text': \" that they don't expect them to be making money in five years because they're not it's the category is this dynamic asset class. The category is not thematic, disruptive tech stocks or X, Y, Z. And and it doesn't work all the time. I mean, it's not and it's not. And and when you layer lots and lots of high fees on it, the markets get more competitive. This is what I mean about.\", 'start': 4857.528, 'end': 4879.448}, {'text': \" you have huge headwinds. You know, you're like, every time you increase, every time you throw in one of these other issues, your probability of generating good returns goes down cumulatively. You don't really have much of a shot. But it is this, it's this, you know, it's trying to basically think about the opportunity to make money in the context of a broader industry.\", 'start': 4879.465, 'end': 4904.947}, {'text': ' And, and, and, and where you can trust the advice, you can trust the advice. So when we were looking at this move from growth to value stocks, I trust a hedge fund who loved when he says values coming back, but he loved growth two years ago, I trust his opinion a lot more than Cathie Woods.', 'start': 4905.875, 'end': 4923.34}, {'text': \" Or someone who's loved value for 50 years. Or GMO or something, who's been in the trade on the wrong side for years and years. If you talk to someone who, I have interviewed folks like this, who their fund is owning oil stocks, don't expect them to be bearish on the price of oil. And that's not their fault. That's who they are. And they went into the industry.\", 'start': 4923.577, 'end': 4947.033}, {'text': \" And people should expect that. They're human beings. So if you go into a job picking hedge funds, you probably like it. You don't want to do some replication robot dog. That might be good for people who want to put something in their portfolio and be happier in 10 years than they would otherwise be. But your job is to pick funds. You don't go back and say, by the way, I don't want to pick funds anymore. I'm going to put it all into this little robot dog of a strategy, this simple, cheap robot ETF.\", 'start': 4947.033, 'end': 4975.788}, {'text': \" and just leave it there. And you should just know that. You should know that if you have your money with an advisor, ask them, so why do I have so many different asset classes? How do you think about, do I really need 25 different line items in this? Do I really need four funds in this category? They may have really great answers for it, but you should feel it's your money and you should feel empowered to ask these kinds of questions.\", 'start': 4975.99, 'end': 5002.923}, {'text': \" I mean, in a sense, our whole business, the other thing that this guy was telling you about kind of like kind of come up with one sentence for a business. He said, in a sense, what you guys have done differently is you just keep asking questions. And when you get these kind of canned responses, if you don't feel that it doesn't make sense to you, you keep asking the same question, which is back to big. I don't get it. What do you mean you don't get it? Like, I don't get it how you think that for the next 70 years, the world is going to look like the past 70 years and that you guys didn't figure this out 10 or 20 years ago.\", 'start': 5003.075, 'end': 5032.572}, {'text': \" Yeah, and the next 70 years is going to be a mirror image of the past 70 years. It's as inane a worldview as it's going to be like the present all the time for the next 70 years. Andrew, I should let you go, but I see you made a lot of interesting points about beta. You said Cathie Wood, don't look to her for alpha, but you said alpha is relative to what? So did the Cathie Wood factor that didn't really exist?\", 'start': 5033.45, 'end': 5057.345}, {'text': \" It did exist, but it was not classified as such. And then now you have the moonshot CTF of unprofitable growth, rich valuation, disruptive tech, however you want to call it. And then also the fact that you said, her clients haven't left her. And again, we're talking about Cathie Wood, it's about the ETF, but the money weighted returns are far, far worse than actually just if you invest.\", 'start': 5057.345, 'end': 5079.923}, {'text': \" They're terrible. They're terrible. Right. But again, that's not her fault. I agree with that. I mean, that's not it's it. Look, it's it was her fault when she was saying buy now because we're going to go up 40 percent a year. Right. That's that's that's crossing a line. That's promising because she was desperately trying to stem outflows. Everything was an argument. But this is what asset managers do. Everything's an argument is why you should remain invested. Back to your point, though, like the the there's a circularity to it. Right. Well, my value guy.\", 'start': 5079.923, 'end': 5107.852}, {'text': ' Do you want your value guide to outperform value stocks?', 'start': 5108.105, 'end': 5111.615}, {'text': \" Or do you want your value guide outperform the S&P 500? Because you think value stocks in it do go well. And there's always a benchmark that performs so badly that your fund will have beaten the index if you choose that index. Exactly. You can jump over it on your knees. Like, oh, I beat the cruise line index in March. You know, great. Look, that just goes back to the dynamics of this industry. Everyone's trying to sell you something. I am trying to, you know, I get on the phone with people, but I'm also very open about it. And because we led with\", 'start': 5111.699, 'end': 5139.745}, {'text': \" with, you know, with our money, we're trying to solve something for ourselves. And so we would invite you to come join us, we think we think we have found better ways of doing a lot of this stuff. But we recognize it's not right for everybody. Yeah, well, what you're doing is very, very interesting. And your final question before I let you go is you do have a peek into what hedge funds are doing on the long short side, as well as managed futures. And you've got a whole process how you do that, we're not gonna have time to get into that. But you do know what what your folks, folks are doing.\", 'start': 5140.117, 'end': 5167.015}, {'text': \" How are hedge funds positioned right now on China? Because I have to say that that is the most unloved, or if it's not, it should be the most unloved asset in the world right now. Because I mean, literally, over the past 10 years, the stock market has been flat as economic growth has skyrocketed. So I know a lot of people who've been, oh, everyone who's been bullish on China has gotten burned. But how are folks positioned now? Because I mean, it could be the ultimate contrarian opportunity. Obviously, the pain could continue for years, if not decades.\", 'start': 5167.015, 'end': 5194.471}, {'text': \" So we don't have good visibility on China positions per se, but I can tell you that it was China was adored three years ago, four years ago by certain particularly tech focused hedge funds. And I think they got slammed. I don't they didn't see the.\", 'start': 5194.471, 'end': 5210.755}, {'text': ' regulatory changes that came and the punishment of the sort of like, like, you know, sharp reversal of this kind of open economic system. Look, my guess is that any contrarian guy worth his salt', 'start': 5211.346, 'end': 5224.745}, {'text': \" is spending a lot of time looking at Chinese companies right now, precisely for the reason you described. There is proverbial blood in the streets for investing in China. It is uninvestable. The macroeconomic situation couldn't look worse. But there will be guys like the guy that I started my career working for, who's trying to find dollars for 25 cents. And he's set up where he doesn't need it to converge tomorrow. He can hold it for 10 years.\", 'start': 5225.487, 'end': 5255.002}, {'text': \" and his clients want him to do that. Yep, China is cheap. Cheap for a reason, perhaps, but it is. Andrew, we'll leave it there. Thanks so much for joining us. People can find you on Twitter at AndrewBeer1, your firm, dbi.co. Thanks for joining, sharing your insights, and thanks, everyone, for watching. Thanks so much. Great talking to you.\", 'start': 5255.187, 'end': 5272.889}, {'text': ' Forward Guidance, the program you just enjoyed, hopefully, can be viewed on YouTube at Blockworks Macro or heard as a podcast on Apple Podcast and Spotify. Episodes are typically released on Apple and Spotify a few hours before they air on YouTube. Please leave a review on Apple Podcast if you feel so inclined. Also, you can get 10% off to Permissionless 2023 and Blockworks Research using code Guidance10. Thanks again and be well.', 'start': 5274.239, 'end': 5298.91}]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "INFO:pytorch_lightning.utilities.migration.utils:Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v2.0.9. To apply the upgrade to your files permanently, run `python -m pytorch_lightning.utilities.upgrade_checkpoint --file ../root/.cache/torch/pyannote/models--pyannote--segmentation/snapshots/c4c8ceafcbb3a7a280c2d357aee9fbc9b0be7f9b/pytorch_model.bin`\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model was trained with pyannote.audio 0.0.1, yours is 2.1.1. Bad things might happen unless you revert pyannote.audio to 0.x.\n",
            "Model was trained with torch 1.10.0+cu102, yours is 2.0.1+cu118. Bad things might happen unless you revert torch to 1.x.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_f5ff730b-04b4-44b3-98b8-aa31566f821c\", \"x1R-VTMknTo.json\", 6920411)"
            ]
          },
          "metadata": {}
        }
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "include_colab_link": true
    },
    "language_info": {
      "name": "python"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "93ceba654c5245d9ab74eef6e4c617f5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_780577efa4634b6fb16c9119cdda58ec",
              "IPY_MODEL_b0dcc88103a8416eab216b362155db83",
              "IPY_MODEL_16bfcd8db8674baca30f989d6f2fb3da"
            ],
            "layout": "IPY_MODEL_0bf38163d89c4f84bed4a1066231b752"
          }
        },
        "780577efa4634b6fb16c9119cdda58ec": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_20f35a634fa5492eb3a77ecc2a3dc44f",
            "placeholder": "​",
            "style": "IPY_MODEL_d00f910737694a0db872991055764be4",
            "value": "Downloading (…)08837e8b/config.json: 100%"
          }
        },
        "b0dcc88103a8416eab216b362155db83": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_b80d79598f0a43bbbcf751d44b977c4f",
            "max": 2796,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_eca6f18b26bf4a3b874962b3df8a3070",
            "value": 2796
          }
        },
        "16bfcd8db8674baca30f989d6f2fb3da": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_f53b929331f94d70a2f0b727948f2e66",
            "placeholder": "​",
            "style": "IPY_MODEL_e8bd1a03cdc943e2bb0ac36905054911",
            "value": " 2.80k/2.80k [00:00&lt;00:00, 84.8kB/s]"
          }
        },
        "0bf38163d89c4f84bed4a1066231b752": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "20f35a634fa5492eb3a77ecc2a3dc44f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d00f910737694a0db872991055764be4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "b80d79598f0a43bbbcf751d44b977c4f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "eca6f18b26bf4a3b874962b3df8a3070": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "f53b929331f94d70a2f0b727948f2e66": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "e8bd1a03cdc943e2bb0ac36905054911": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "9b5c9787f8d346caaff7e8f536704f4b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_b93529f8a3bf4eac8d0de82ee8ee8325",
              "IPY_MODEL_322411b766ae4dcca07aa80f3b9cc308",
              "IPY_MODEL_ce1610b9833d43d985d671dd36faa7ea"
            ],
            "layout": "IPY_MODEL_cb5e7a767c57466e9f5c1a26b56d5853"
          }
        },
        "b93529f8a3bf4eac8d0de82ee8ee8325": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5cdfaeb9f2dd403b9d1cf844e9695dae",
            "placeholder": "​",
            "style": "IPY_MODEL_add074b8ec6143d28dc24de836656941",
            "value": "Downloading (…)37e8b/vocabulary.txt: 100%"
          }
        },
        "322411b766ae4dcca07aa80f3b9cc308": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d6559426c08045ab867bf849204124d8",
            "max": 459861,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_ebda834cc20a4e7381c2c0665b5dc33b",
            "value": 459861
          }
        },
        "ce1610b9833d43d985d671dd36faa7ea": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_33d0015ff67d4162902af4695b35ee53",
            "placeholder": "​",
            "style": "IPY_MODEL_104fa4bc207c4ccd82f2cffbabcc916c",
            "value": " 460k/460k [00:00&lt;00:00, 5.27MB/s]"
          }
        },
        "cb5e7a767c57466e9f5c1a26b56d5853": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "5cdfaeb9f2dd403b9d1cf844e9695dae": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "add074b8ec6143d28dc24de836656941": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "d6559426c08045ab867bf849204124d8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "ebda834cc20a4e7381c2c0665b5dc33b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "33d0015ff67d4162902af4695b35ee53": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "104fa4bc207c4ccd82f2cffbabcc916c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "0625c2337af64c6fa2e33d58a614e0e8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_135589fbf8ec402aad213b5ce38f82df",
              "IPY_MODEL_76688a3bcc4443e583e841f6586544a5",
              "IPY_MODEL_f2fa15344321437bb2de6fcae3785b4f"
            ],
            "layout": "IPY_MODEL_ed86c9845d4d40708d3dc8ae8c11d52b"
          }
        },
        "135589fbf8ec402aad213b5ce38f82df": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a55c3d810b49482d8487668eeaba72ee",
            "placeholder": "​",
            "style": "IPY_MODEL_1960d9e9f48045f0ba64ee34b8b7927d",
            "value": "Downloading model.bin: 100%"
          }
        },
        "76688a3bcc4443e583e841f6586544a5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d2e2dd6afb8d41b782e149cb2b4defa5",
            "max": 3086912962,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_d1df5de8540345feb757b6256b6d0128",
            "value": 3086912962
          }
        },
        "f2fa15344321437bb2de6fcae3785b4f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_938ac00af36b4d86b04e406974b2c00d",
            "placeholder": "​",
            "style": "IPY_MODEL_6abaf16e6cd341e2900cb28c82e23f65",
            "value": " 3.09G/3.09G [00:12&lt;00:00, 264MB/s]"
          }
        },
        "ed86c9845d4d40708d3dc8ae8c11d52b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a55c3d810b49482d8487668eeaba72ee": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "1960d9e9f48045f0ba64ee34b8b7927d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "d2e2dd6afb8d41b782e149cb2b4defa5": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d1df5de8540345feb757b6256b6d0128": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "938ac00af36b4d86b04e406974b2c00d": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6abaf16e6cd341e2900cb28c82e23f65": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "f55c189b5c9e449696fa9e7a09ccabfb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_a1348e8aa1b047608ae14b214dc6d978",
              "IPY_MODEL_5321f147d82843d9a21d74e75153d644",
              "IPY_MODEL_ffa9403b2d984ee7b810a7dee86ce21d"
            ],
            "layout": "IPY_MODEL_2e3fb4ce3bf4482db12ddc6102f56ad4"
          }
        },
        "a1348e8aa1b047608ae14b214dc6d978": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_86d08e9b45c5430f9bd35bfbad249ad5",
            "placeholder": "​",
            "style": "IPY_MODEL_812ab7be10bc44b3a948aa14ebab46b2",
            "value": "Downloading (…)37e8b/tokenizer.json: 100%"
          }
        },
        "5321f147d82843d9a21d74e75153d644": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_7b38c06843b94dc696d95d0ac9af6ae8",
            "max": 2203239,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_d25a36bb87ce41ae948b7d3d204a6c95",
            "value": 2203239
          }
        },
        "ffa9403b2d984ee7b810a7dee86ce21d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_47161a11268f4b67923e65d44b2f15f0",
            "placeholder": "​",
            "style": "IPY_MODEL_8f4fd993adfd44359250c5a6eed3885f",
            "value": " 2.20M/2.20M [00:00&lt;00:00, 4.93MB/s]"
          }
        },
        "2e3fb4ce3bf4482db12ddc6102f56ad4": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "86d08e9b45c5430f9bd35bfbad249ad5": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "812ab7be10bc44b3a948aa14ebab46b2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "7b38c06843b94dc696d95d0ac9af6ae8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d25a36bb87ce41ae948b7d3d204a6c95": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "47161a11268f4b67923e65d44b2f15f0": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8f4fd993adfd44359250c5a6eed3885f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}