doc/source/models/overview.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[source](../api/alibi_detect.models.rst)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Overview\n",
    "\n",
    "Models and/or building blocks that can be useful outside of outlier, adversarial or drift detection can be found under `alibi_detect.models`. Main implementations:\n",
    "\n",
    "- Variational Autoencoder: `alibi_detect.models.autoencoder.VAE`\n",
    "\n",
    "- Sequence-to-sequence model: `alibi_detect.models.autoencoder.Seq2Seq`\n",
    "\n",
    "- ResNet: `alibi_detect.models.resnet`\n",
    "\n",
    "Pre-trained ResNet-20/32/44 models on CIFAR-10 can be found on our [Google Cloud Bucket](https://console.cloud.google.com/storage/browser/seldon-models/alibi-detect/classifier/cifar10/?organizationId=156002945562&project=seldon-pub) and can be fetched as follows:\n",
    "\n",
    "```python\n",
    "from alibi_detect.utils.fetching import fetch_tf_model\n",
    "\n",
    "model = fetch_tf_model('cifar10', 'resnet32')\n",
    "```"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:detect] *",
   "language": "python",
   "name": "conda-env-detect-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}