{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "23db9840",
   "metadata": {},
   "source": [
    "Hi Georg,\n",
    "\n",
    "Thanks for reaching out. I actually graduated from MIT recently so it’s cool to hear that they’ll be learning about this!\n",
    "I’ve attached the notebook we used for running the benchmarks (heavily adopted from pyctcdecode’s performance eval tutorial), and here’s a tutorial for getting started with the TorchAudio decoder as well. Feel free to send over any feedback or suggestions for the decoder, always happy to hear about potential ways to make it even better!\n",
    "\n",
    "Best,\n",
    "Caroline\n",
    "\n",
    "https://github.com/kensho-technologies/pyctcdecode/blob/main/tutorials/03_eval_performance.ipynb\n",
    "https://pytorch.org/audio/main/tutorials/asr_inference_with_ctc_decoder_tutorial.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6d5c4dc",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "980a108d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e7b0aab",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bea7083b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}