Autoencoders_VAE_MNIST_JaxvPyTorch.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "692f5170",
   "metadata": {},
   "source": [
    "# I. Flax implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b13be0e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import jax\n",
    "from jax import lax, random, numpy as jnp\n",
    "# No function to convert PIL image to jax array : intermediary conversion to numpy array necessary\n",
    "import numpy as np\n",
    "\n",
    "import flax\n",
    "from flax import struct, linen as nn\n",
    "from flax.training import train_state\n",
    "import optax\n",
    "from clu import metrics\n",
    "\n",
    "# No native dataloading mechanism in Jax\n",
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision.datasets import MNIST\n",
    "\n",
    "from typing import Sequence\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm.notebook import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fade0ff1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Globales\n",
    "batch_size = 256\n",
    "workers = 8\n",
    "learning_rate = 0.001\n",
    "epochs = 100\n",
    "seed = 654321"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08cc6e62",
   "metadata": {},
   "source": [
    "## 1.1. Data preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "caed90f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def flatten_transform(x):\n",
    "    \"\"\" Scales values from [0, 255] to [0, 1] and flattens the 2D images (28x28 -> 784) \"\"\"\n",
    "    return np.ravel(np.array(x, dtype=jnp.float32) / 255.)\n",
    "\n",
    "def numpy_collate(batch):\n",
    "    \"\"\" Stack elements in batches of numpy arrays instead of Torch tensors \"\"\"\n",
    "    transposed_data = list(zip(*batch))\n",
    "    imgs = np.stack(transposed_data[0])\n",
    "    labels = np.array(transposed_data[1])\n",
    "\n",
    "    return imgs, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d396061",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 10000)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainset = MNIST(root=\"./data\", train=True, download=True, transform=flatten_transform)\n",
    "testset = MNIST(root=\"./data\", train=False, download=True, transform=flatten_transform)\n",
    "trainloader = DataLoader(trainset, batch_size, shuffle=True, num_workers=workers, collate_fn=numpy_collate, drop_last=True)\n",
    "testloader = DataLoader(testset, batch_size, shuffle=False, num_workers=workers, collate_fn=numpy_collate, drop_last=True)\n",
    "len(trainset), len(testset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f117ebf2",
   "metadata": {},
   "source": [
    "Display of the first 10 images of the trainloader (not the first 10 of the dataset since shuffle=True) to confirm that the loaders have been correctly created"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3c971813",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# For the pytorch dataloader \n",
    "# https://discuss.pytorch.org/t/reproducibility-dataloader-shuffle-true-using-seeds/173836/2\n",
    "torch.manual_seed(seed)\n",
    "imgs, _ = next(iter(trainloader))\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(1, 10, i + 1)\n",
    "    img = imgs[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c61dc52b",
   "metadata": {},
   "source": [
    "## 1.2. Vanilla autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d20cb6f",
   "metadata": {},
   "source": [
    "### 1.2.1. Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e892dab7",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Encoder(nn.Module):\n",
    "    \"\"\" A simple MLP encoder \"\"\"\n",
    "    latents: Sequence[int]\n",
    "    \n",
    "    @nn.compact\n",
    "    def __call__(self, x):\n",
    "        for i, n in enumerate(self.latents):\n",
    "            x = nn.Dense(features=n)(x)\n",
    "            if i != len(self.latents) - 1:\n",
    "                x = nn.relu(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e53580b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Decoder(nn.Module):\n",
    "    \"\"\" A simple MLP decoder \"\"\"\n",
    "    latents: Sequence[int]\n",
    "\n",
    "    @nn.compact\n",
    "    def __call__(self, x):\n",
    "        for i, n in enumerate(self.latents):\n",
    "            x = nn.Dense(features=n)(x)\n",
    "            # No activation function for last layer\n",
    "            # cf. use of sigmoid_binary_cross_entropy as loss function\n",
    "            if i != len(self.latents) - 1:\n",
    "                x = nn.relu(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "64c90087",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Autoencoder(nn.Module):\n",
    "    \"\"\" Full autoencoder model \"\"\"\n",
    "    latents: Sequence[int]\n",
    "\n",
    "    def setup(self):\n",
    "        self.encoder = Encoder(self.latents)\n",
    "        self.decoder = Decoder(([784] + list(self.latents[:-1]))[::-1])\n",
    "\n",
    "    def __call__(self, x):\n",
    "        z = self.encoder(x)\n",
    "        return self.decoder(z)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a53e495a",
   "metadata": {},
   "source": [
    "Full model inspection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9376299f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b[3m                              Autoencoder Summary                               \u001b[0m\n",
      "┏━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
      "┃\u001b[1m \u001b[0m\u001b[1mpath         \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mmodule     \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1minputs       \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1moutputs       \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mparams       \u001b[0m\u001b[1m \u001b[0m┃\n",
      "┡━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
      "│               │ Autoencoder │ \u001b[2mfloat32\u001b[0m[1,78… │ \u001b[2mfloat32\u001b[0m[1,784] │               │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ encoder       │ Encoder     │ \u001b[2mfloat32\u001b[0m[1,78… │ \u001b[2mfloat32\u001b[0m[1,64]  │               │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ encoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,78… │ \u001b[2mfloat32\u001b[0m[1,512] │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[512]  │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[784,… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m401,920 \u001b[0m\u001b[1;2m(1.6 \u001b[0m │\n",
      "│               │             │               │                │ \u001b[1;2mMB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ encoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,51… │ \u001b[2mfloat32\u001b[0m[1,128] │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[128]  │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[512,… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m65,664 \u001b[0m\u001b[1;2m(262.7\u001b[0m │\n",
      "│               │             │               │                │ \u001b[1;2mKB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ encoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,12… │ \u001b[2mfloat32\u001b[0m[1,64]  │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[64]   │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[128,… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m8,256 \u001b[0m\u001b[1;2m(33.0 \u001b[0m  │\n",
      "│               │             │               │                │ \u001b[1;2mKB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ decoder       │ Decoder     │ \u001b[2mfloat32\u001b[0m[1,64] │ \u001b[2mfloat32\u001b[0m[1,784] │               │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ decoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,64] │ \u001b[2mfloat32\u001b[0m[1,128] │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[128]  │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[64,1… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m8,320 \u001b[0m\u001b[1;2m(33.3 \u001b[0m  │\n",
      "│               │             │               │                │ \u001b[1;2mKB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ decoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,12… │ \u001b[2mfloat32\u001b[0m[1,512] │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[512]  │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[128,… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m66,048 \u001b[0m\u001b[1;2m(264.2\u001b[0m │\n",
      "│               │             │               │                │ \u001b[1;2mKB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│ decoder/Dens… │ Dense       │ \u001b[2mfloat32\u001b[0m[1,51… │ \u001b[2mfloat32\u001b[0m[1,784] │ bias:         │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[784]  │\n",
      "│               │             │               │                │ kernel:       │\n",
      "│               │             │               │                │ \u001b[2mfloat32\u001b[0m[512,… │\n",
      "│               │             │               │                │               │\n",
      "│               │             │               │                │ \u001b[1m402,192 \u001b[0m\u001b[1;2m(1.6 \u001b[0m │\n",
      "│               │             │               │                │ \u001b[1;2mMB)\u001b[0m           │\n",
      "├───────────────┼─────────────┼───────────────┼────────────────┼───────────────┤\n",
      "│\u001b[1m \u001b[0m\u001b[1m             \u001b[0m\u001b[1m \u001b[0m│\u001b[1m \u001b[0m\u001b[1m           \u001b[0m\u001b[1m \u001b[0m│\u001b[1m \u001b[0m\u001b[1m             \u001b[0m\u001b[1m \u001b[0m│\u001b[1m \u001b[0m\u001b[1m         Total\u001b[0m\u001b[1m \u001b[0m│\u001b[1m \u001b[0m\u001b[1m952,400 \u001b[0m\u001b[1;2m(3.8 \u001b[0m\u001b[1m \u001b[0m│\n",
      "│\u001b[1m               \u001b[0m│\u001b[1m             \u001b[0m│\u001b[1m               \u001b[0m│\u001b[1m                \u001b[0m│\u001b[1m \u001b[0m\u001b[1;2mMB)\u001b[0m\u001b[1m          \u001b[0m\u001b[1m \u001b[0m│\n",
      "└───────────────┴─────────────┴───────────────┴────────────────┴───────────────┘\n",
      "\u001b[1m                                                                                \u001b[0m\n",
      "\u001b[1m                       Total Parameters: 952,400 \u001b[0m\u001b[1;2m(3.8 MB)\u001b[0m\u001b[1m                       \u001b[0m\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "autoencoder = Autoencoder([512, 128, 64])\n",
    "print(autoencoder.tabulate(random.PRNGKey(seed), jnp.ones((1, 784))))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a7626ce",
   "metadata": {},
   "source": [
    "### 1.2.2. Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3cb207d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://flax.readthedocs.io/en/latest/getting_started.html#create-a-trainstate\n",
    "@struct.dataclass\n",
    "class Metrics(metrics.Collection):\n",
    "    loss: metrics.Average.from_output('loss')\n",
    "\n",
    "\n",
    "class TrainState(train_state.TrainState):\n",
    "    metrics: Metrics\n",
    "\n",
    "\n",
    "def create_train_state(model, rng, learning_rate):\n",
    "    \"\"\" Instanciate the state outside of the training loop \"\"\"\n",
    "    params = model.init(rng, jnp.ones([1, 784]))['params']\n",
    "    opti = optax.adam(learning_rate)\n",
    "    return TrainState.create(\n",
    "        apply_fn=model.apply, params=params, tx=opti,\n",
    "        metrics=Metrics.empty())\n",
    "\n",
    "\n",
    "@jax.jit\n",
    "def train_step(state, batch):\n",
    "    \"\"\" Train for a single step \"\"\"\n",
    "    def loss_fn(params):\n",
    "        logits = state.apply_fn({'params': params}, batch[0])\n",
    "        loss = optax.sigmoid_binary_cross_entropy(\n",
    "            logits=logits, labels=batch[0]).mean()\n",
    "        return loss, logits\n",
    "    \n",
    "    # Update parameters with gradient descent\n",
    "    grad_fn = jax.value_and_grad(loss_fn, has_aux=True)\n",
    "    (loss, logits), grads = grad_fn(state.params)\n",
    "    state = state.apply_gradients(grads=grads)\n",
    "    \n",
    "    # Update loss, stored in the state object\n",
    "    metric_updates = state.metrics.single_from_model_output(\n",
    "        logits=logits, labels=batch[0], loss=loss)\n",
    "    metrics = state.metrics.merge(metric_updates)\n",
    "    state = state.replace(metrics=metrics)\n",
    "    \n",
    "    return state\n",
    "\n",
    "\n",
    "@jax.jit\n",
    "def eval_step(state, batch):\n",
    "    \"\"\" Computes the metric on the test batch (code already included in train_step for train batch) \"\"\"\n",
    "    logits = state.apply_fn({'params': state.params}, batch[0])\n",
    "    loss = optax.sigmoid_binary_cross_entropy(\n",
    "        logits=logits, labels=batch[0]).mean()\n",
    "    metric_updates = state.metrics.single_from_model_output(\n",
    "        logits=logits, labels=batch[0], loss=loss)\n",
    "    metrics = state.metrics.merge(metric_updates)\n",
    "    state = state.replace(metrics=metrics)\n",
    "    return state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cc099a48",
   "metadata": {},
   "outputs": [],
   "source": [
    "init_rng = random.PRNGKey(seed)\n",
    "state = create_train_state(autoencoder, init_rng, learning_rate)\n",
    "del init_rng"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0a52c137",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "82aca40962ca4cc5adc3d9270015676d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 8s, sys: 36.9 s, total: 1min 45s\n",
      "Wall time: 2min 13s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    for batch in trainloader:\n",
    "        state = train_step(state, batch)\n",
    "    train_loss.append(state.metrics.compute()['loss'].item())\n",
    "    # Reset train_metrics for next training epoch\n",
    "    state = state.replace(metrics=state.metrics.empty())\n",
    "    \n",
    "    # Compute metrics on the test set after each training epoch\n",
    "    test_state = state\n",
    "    for batch in testloader:\n",
    "        test_state = eval_step(test_state, batch)\n",
    "    test_loss.append(test_state.metrics.compute()['loss'].item())\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84e8d091",
   "metadata": {},
   "source": [
    "### 1.2.3. Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a0faf86c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(np.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(np.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Evolution of the losses\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "80606d25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "rec = nn.sigmoid(state.apply_fn({'params': state.params}, imgs))\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = imgs[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = rec[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0aaa0db8",
   "metadata": {},
   "source": [
    "The results are quite good, but the dataset is really simple (and the network is probably overparametrized for such an easy task). <br>\n",
    "Let's set the latent space dimension to 2 to visualize the representations learned by the neural network. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de0fcacf",
   "metadata": {},
   "source": [
    "### 1.2.4. Latent space visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fe2f6a51",
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder = Autoencoder([512, 128, 64, 2])\n",
    "init_rng = random.PRNGKey(seed)\n",
    "state = create_train_state(autoencoder, init_rng, learning_rate)\n",
    "del init_rng"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "452298cd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0ebdb49b16294bbab5fa4704c6daee21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 20s, sys: 40.4 s, total: 2min\n",
      "Wall time: 2min 20s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    for batch in trainloader:\n",
    "        state = train_step(state, batch)\n",
    "    train_loss.append(state.metrics.compute()['loss'].item())\n",
    "    # Reset train_metrics for next training epoch\n",
    "    state = state.replace(metrics=state.metrics.empty())\n",
    "    \n",
    "    # Compute metrics on the test set after each training epoch\n",
    "    test_state = state\n",
    "    for batch in testloader:\n",
    "        test_state = eval_step(test_state, batch)\n",
    "    test_loss.append(test_state.metrics.compute()['loss'].item())\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "34206ec7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "rec = nn.sigmoid(state.apply_fn({'params': state.params}, imgs))\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = imgs[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = rec[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f047ef3d",
   "metadata": {},
   "source": [
    "As could be expected, the results are more blurry, but still decent for a model with only 2 neurons in the latent space !"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "de87003d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(np.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(np.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Evolution of the losses (latent dimension = 2)\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35190996",
   "metadata": {},
   "source": [
    "Encoding of the train set into latent space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "574dcb01",
   "metadata": {},
   "outputs": [],
   "source": [
    "encoder = Encoder([512, 128, 64, 2])\n",
    "init_rng = random.PRNGKey(seed)\n",
    "enc_state = create_train_state(encoder, init_rng, learning_rate)\n",
    "del init_rng\n",
    "\n",
    "embeddings, labels = [], []\n",
    "full_trainloader = DataLoader(trainset, batch_size, shuffle=True, num_workers=workers, collate_fn=numpy_collate)\n",
    "torch.manual_seed(seed)\n",
    "for batch in full_trainloader:\n",
    "    embeddings.append(enc_state.apply_fn({'params': state.params['encoder']}, batch[0]))\n",
    "    labels.append(batch[1])\n",
    "embeddings = jnp.vstack(embeddings)\n",
    "labels = jnp.concatenate(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9b41cd45",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9, 6))\n",
    "plt.scatter(embeddings[:, 0], embeddings[:, 1], c=labels, cmap='Spectral', s=3)\n",
    "plt.colorbar(boundaries=np.arange(11)-0.5).set_ticks(np.arange(10))\n",
    "plt.xlabel(\"First latent dimension\")\n",
    "plt.ylabel(\"Second latent dimension\")\n",
    "plt.title('Visualisation of the latent space of an autoencoder trained on MNIST');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "785eac94",
   "metadata": {},
   "source": [
    "Without applying any constraint, we can still observe some form of clustering : most numbers are localized into a specific zone of the 2D latent space. This only works because MNIST is a relatively simple dataset, with low variance within each class. <br>\n",
    "However, this representation helps understand the limitations of autoencoders as generative models :\n",
    " * The latent space is not continuous\n",
    " * The values on the different dimensions are not bounded  : scale between -30 and 60 along the first dimension, and between -40 and 40 for the second dimension\n",
    " \n",
    "This makes it difficult to generate new data, as randomly sampling the latent space will most likely result in sampling from unmapped space, i.e. generating an image likely unrelated to the training set.\n",
    "\n",
    "A few examples are shown below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cbd1d6de",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoder = Decoder([64, 128, 512, 784])\n",
    "init_rng = random.PRNGKey(seed)\n",
    "dec_state = create_train_state(decoder, init_rng, learning_rate)\n",
    "del init_rng\n",
    "\n",
    "samples = jnp.array([[-20, 0], [-10, 10], [0, 20], [0, 50], [10, 200], [20, 30], [25, 25], [100, 25], [1000, -1000]])\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i, sample in enumerate(samples):\n",
    "    ax = plt.subplot(1, samples.shape[0], i + 1)\n",
    "    img = nn.sigmoid(dec_state.apply_fn({'params': state.params['decoder']}, sample))\n",
    "    ax.imshow(img.reshape((28, 28)), cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(f\"Sample {sample}\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae003dca",
   "metadata": {},
   "source": [
    "The dataset being rather simple, the generated images for coordinates outside of the learned ranges still produce somehow recognisable numbers."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7d7dc72",
   "metadata": {},
   "source": [
    "## 1.3. VAE"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da590515",
   "metadata": {},
   "source": [
    "### 1.3.1. Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cb535b5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "class VEncoder(nn.Module):\n",
    "    \"\"\" Encoder of the VAE \"\"\"\n",
    "    latents: Sequence[int]\n",
    "    \n",
    "    @nn.compact\n",
    "    def __call__(self, x):\n",
    "        for n in self.latents[:-1]:\n",
    "            x = nn.Dense(features=n)(x)\n",
    "            x = nn.relu(x)\n",
    "        mean_x = nn.Dense(self.latents[-1])(x)\n",
    "        logvar_x = nn.Dense(self.latents[-1])(x)\n",
    "        return mean_x, logvar_x\n",
    "\n",
    "\n",
    "class VDecoder(nn.Module):\n",
    "    \"\"\" Decoder of the VAE \"\"\"\n",
    "    latents: Sequence[int]\n",
    "\n",
    "    @nn.compact\n",
    "    def __call__(self, x):\n",
    "        for i, n in enumerate(self.latents):\n",
    "            x = nn.Dense(features=n)(x)\n",
    "            # No activation function for last layer\n",
    "            # cf. use of sigmoid_binary_cross_entropy as loss function\n",
    "            if i != len(self.latents) - 1:\n",
    "                x = nn.relu(x)\n",
    "        return x\n",
    "\n",
    "\n",
    "def sample(rng, mean, logvar):\n",
    "    \"\"\" Sample latent vector from mean and log variance \"\"\"\n",
    "    std = jnp.exp(0.5 * logvar) # 0.5 for root square, exp to cancel log\n",
    "    eps = random.normal(rng, logvar.shape)\n",
    "    return mean + eps * std\n",
    "\n",
    "\n",
    "class VAE(nn.Module):\n",
    "    \"\"\" Full VAE model \"\"\"\n",
    "    latents: Sequence[int]\n",
    "\n",
    "    def setup(self):\n",
    "        self.encoder = VEncoder(self.latents)\n",
    "        self.decoder = VDecoder(([784] + list(self.latents[:-1]))[::-1])\n",
    "\n",
    "    def __call__(self, x, z_rng):\n",
    "        mean, logvar = self.encoder(x)\n",
    "        z = sample(z_rng, mean, logvar)\n",
    "        x_prime = self.decoder(z)\n",
    "        return self.decoder(z), mean, logvar"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d258d449",
   "metadata": {},
   "source": [
    "### 1.3.2. Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1f8dbd29",
   "metadata": {},
   "outputs": [],
   "source": [
    "@struct.dataclass\n",
    "class Metrics(metrics.Collection):\n",
    "    loss: metrics.Average.from_output('loss')\n",
    "\n",
    "\n",
    "class TrainState(train_state.TrainState):\n",
    "    metrics: Metrics\n",
    "\n",
    "\n",
    "def create_train_state(model, rng, learning_rate, z_rng):\n",
    "    \"\"\" Instanciate the state outside of the training loop \"\"\"\n",
    "    params = model.init(rng, jnp.ones([1, 784]), z_rng)['params']\n",
    "    opti = optax.adam(learning_rate)\n",
    "    return TrainState.create(\n",
    "        apply_fn=model.apply, params=params, tx=opti,\n",
    "        metrics=Metrics.empty())\n",
    "\n",
    "\n",
    "def compute_loss(x, x_rec, mean, logvar):\n",
    "    \"\"\" VAE loss is binary cross entropy (reconstruction loss) + KL divergence (regularization of latent space)\"\"\"\n",
    "    bce = optax.sigmoid_binary_cross_entropy(logits=x_rec, labels=x).sum()\n",
    "    kl = -0.5 * jnp.sum(1 + logvar - jnp.square(mean) - jnp.exp(logvar))\n",
    "    return bce + kl\n",
    "\n",
    "\n",
    "@jax.jit\n",
    "def train_step(state, batch, z_rng):\n",
    "    \"\"\" Train for a single step \"\"\"\n",
    "    def loss_fn(params):\n",
    "        logits, mean, logvar = state.apply_fn({'params': params}, batch[0], z_rng)\n",
    "        loss = compute_loss(batch[0], logits, mean, logvar)\n",
    "        return loss, logits\n",
    "    \n",
    "    # Update parameters with gradient descent\n",
    "    grad_fn = jax.value_and_grad(loss_fn, has_aux=True)\n",
    "    (loss, logits), grads = grad_fn(state.params)\n",
    "    state = state.apply_gradients(grads=grads)\n",
    "    \n",
    "    # Update loss, stored in the state object\n",
    "    metric_updates = state.metrics.single_from_model_output(\n",
    "        logits=logits, labels=batch[0], loss=loss)\n",
    "    metrics = state.metrics.merge(metric_updates)\n",
    "    state = state.replace(metrics=metrics)\n",
    "    \n",
    "    return state\n",
    "\n",
    "\n",
    "@jax.jit\n",
    "def eval_step(state, batch, z_rng):\n",
    "    \"\"\" Computes the metric on the test batch (code already included in train_step for train batch) \"\"\"\n",
    "    logits, mean, logvar = state.apply_fn({'params': state.params}, batch[0], z_rng)\n",
    "    loss = compute_loss(batch[0], logits, mean, logvar)\n",
    "    metric_updates = state.metrics.single_from_model_output(\n",
    "        logits=logits, labels=batch[0], loss=loss)\n",
    "    metrics = state.metrics.merge(metric_updates)\n",
    "    state = state.replace(metrics=metrics)\n",
    "    return state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "89a7e296",
   "metadata": {},
   "outputs": [],
   "source": [
    "vae = VAE([512, 128, 64])\n",
    "init_rng = random.PRNGKey(seed)\n",
    "init_rng, z_init_rng, z_rng, eval_rng = random.split(init_rng, 4)\n",
    "state = create_train_state(vae, init_rng, learning_rate, z_init_rng)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b145e65e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fd90045de2b74613973d32605b5bb7b4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 35s, sys: 49.3 s, total: 2min 24s\n",
      "Wall time: 2min 33s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    for batch in trainloader:\n",
    "        z_rng, z_key = random.split(z_rng)\n",
    "        state = train_step(state, batch, z_key)\n",
    "    train_loss.append(state.metrics.compute()['loss'].item())\n",
    "    # Reset train_metrics for next training epoch\n",
    "    state = state.replace(metrics=state.metrics.empty())\n",
    "    \n",
    "    # Compute metrics on the test set after each training epoch\n",
    "    test_state = state\n",
    "    for batch in testloader:\n",
    "        eval_rng, eval_key = random.split(eval_rng)\n",
    "        test_state = eval_step(test_state, batch, eval_key)\n",
    "    test_loss.append(test_state.metrics.compute()['loss'].item())\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "191c696e",
   "metadata": {},
   "source": [
    "### 1.3.3. Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "98b868f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(np.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(np.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Learning curves for the VAE\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca955ba5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "eval_rng, eval_key = random.split(eval_rng)\n",
    "rec, _, _ = state.apply_fn({'params': state.params}, imgs, eval_key)\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = nn.sigmoid(imgs[i]).reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = nn.sigmoid(rec[i]).reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfc426a5",
   "metadata": {},
   "source": [
    "The quality of images obtained seems on par with the one of images generated with a vanilla autoencoder. The losses profile is also very similar."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "490e4099",
   "metadata": {},
   "source": [
    "### 1.3.4. Latent space visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "009c42ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "vae = VAE([512, 128, 64, 2])\n",
    "init_rng = random.PRNGKey(seed)\n",
    "init_rng, z_init_rng, z_rng, eval_rng = random.split(init_rng, 4)\n",
    "state = create_train_state(vae, init_rng, learning_rate, z_init_rng)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "db71f7fb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "70681db14d8f4c9d8669e48cc3cff202",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 43s, sys: 49.7 s, total: 2min 33s\n",
      "Wall time: 2min 34s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    for batch in trainloader:\n",
    "        z_rng, z_key = random.split(z_rng)\n",
    "        state = train_step(state, batch, z_key)\n",
    "    train_loss.append(state.metrics.compute()['loss'].item())\n",
    "    # Reset train_metrics for next training epoch\n",
    "    state = state.replace(metrics=state.metrics.empty())\n",
    "    \n",
    "    # Compute metrics on the test set after each training epoch\n",
    "    test_state = state\n",
    "    for batch in testloader:\n",
    "        eval_rng, eval_key = random.split(eval_rng)\n",
    "        test_state = eval_step(test_state, batch, eval_key)\n",
    "    test_loss.append(test_state.metrics.compute()['loss'].item())\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1eeac7c8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(np.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(np.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Learning curves for the VAE (latent dim = 2)\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84884cf8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "eval_rng, eval_key = random.split(eval_rng)\n",
    "rec, _, _ = state.apply_fn({'params': state.params}, imgs, eval_key)\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = nn.sigmoid(imgs[i]).reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = nn.sigmoid(rec[i]).reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9199298",
   "metadata": {},
   "source": [
    "As with the vanilla autoencoder, images are blurrier but still decent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "41891898",
   "metadata": {},
   "outputs": [],
   "source": [
    "vencoder = VEncoder([512, 128, 64, 2])\n",
    "venc_params = vencoder.init(init_rng, jnp.ones([1, 784]))['params']\n",
    "opti = optax.adam(learning_rate)\n",
    "venc_state = TrainState.create(apply_fn=vencoder.apply, params=venc_params,\n",
    "                               tx=opti, metrics=Metrics.empty())\n",
    "\n",
    "embeddings, labels = [], []\n",
    "full_trainloader = DataLoader(trainset, batch_size, shuffle=True, num_workers=workers, collate_fn=numpy_collate)\n",
    "torch.manual_seed(seed)\n",
    "for batch in full_trainloader:\n",
    "    mean, logvar = venc_state.apply_fn({'params': state.params['encoder']}, batch[0])\n",
    "    embeddings.append(mean)\n",
    "    labels.append(batch[1])\n",
    "embeddings = jnp.vstack(embeddings)\n",
    "labels = jnp.concatenate(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6576065f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9, 6))\n",
    "plt.scatter(embeddings[:, 0], embeddings[:, 1], c=labels, cmap='Spectral', s=3)\n",
    "plt.colorbar(boundaries=np.arange(11)-0.5).set_ticks(np.arange(10))\n",
    "plt.xlabel(\"First latent dimension\")\n",
    "plt.ylabel(\"Second latent dimension\")\n",
    "plt.title('Visualisation of the latent space of a VAE trained on MNIST');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a008b021",
   "metadata": {},
   "source": [
    "Since we are only plotting the coordinates of the mean, the differences constated here, in comparison to the latent space of the autoencoder, are only due to the influence of the KL term in the loss (same architecture and loss otherwise). <br>\n",
    "Due to the KL, the clusters obtained are closer to a gaussian distribution of standard deviation 1. As a consequence, the points are less spread than for the vanilla autoencoder (dimension values between -4 and 4 this time) : the latent space is continuous in the center."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d704131d",
   "metadata": {},
   "source": [
    "# II. PyTorch implementation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00194557",
   "metadata": {},
   "source": [
    "In order to compare the performances, the same models are implemented in PyTorch below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "755ed015",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn # beware, same alias as flax.linen\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import DataLoader\n",
    "import torchvision.transforms as T\n",
    "from torchvision.datasets import MNIST\n",
    "from torchinfo import summary\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm.notebook import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "78778df2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Globales\n",
    "batch_size = 256\n",
    "workers = 8\n",
    "learning_rate = 0.001\n",
    "epochs = 100\n",
    "seed = 654321\n",
    "device = torch.device(\"cuda\")\n",
    "torch.cuda.manual_seed_all(seed)\n",
    "torch.set_float32_matmul_precision('high')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ed2ce4ed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 10000, 234, 39)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainset = MNIST(root=\"./data\", train=True, download=True,\n",
    "                 transform=T.Compose([T.ToTensor(), torch.flatten]))\n",
    "testset = MNIST(root=\"./data\", train=False, download=True,\n",
    "                transform=T.Compose([T.ToTensor(), torch.flatten]))\n",
    "trainloader = DataLoader(trainset, batch_size, shuffle=True, num_workers=workers, drop_last=True)\n",
    "testloader = DataLoader(testset, batch_size, shuffle=False, num_workers=workers, drop_last=True)\n",
    "len(trainset), len(testset), len(trainloader), len(testloader)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b07ac882",
   "metadata": {},
   "source": [
    "## 2.1. Vanilla autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e7e7922",
   "metadata": {},
   "source": [
    "### 2.1.1. Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba280a1f",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Autoencoder(nn.Module):\n",
    "    def __init__(self, latents):\n",
    "        \"\"\" Autoencoder constructor \"\"\"\n",
    "        super(Autoencoder, self).__init__()\n",
    "        self.latents = [784] + latents\n",
    "        self.encoder = nn.Sequential()\n",
    "        self.decoder = nn.Sequential()\n",
    "        for i, l in enumerate(self.latents[:-1]):\n",
    "            self.encoder.append(nn.Linear(l, self.latents[i+1]))\n",
    "            if i != len(self.latents) - 2:\n",
    "                x = nn.ReLU()\n",
    "        for i, l in enumerate(self.latents[::-1][:-1]):\n",
    "            self.decoder.append(nn.Linear(l, self.latents[::-1][i+1]))\n",
    "            if i != len(self.latents) - 2:\n",
    "                x = nn.ReLU()\n",
    "\n",
    "    def forward(self, x):\n",
    "        z = self.encoder(x)\n",
    "        return self.decoder(z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "641cc8ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "=================================================================\n",
       "Layer (type:depth-idx)                   Param #\n",
       "=================================================================\n",
       "OptimizedModule                          --\n",
       "├─Autoencoder: 1-1                       --\n",
       "│    └─Sequential: 2-1                   --\n",
       "│    │    └─Linear: 3-1                  401,920\n",
       "│    │    └─Linear: 3-2                  65,664\n",
       "│    │    └─Linear: 3-3                  8,256\n",
       "│    └─Sequential: 2-2                   --\n",
       "│    │    └─Linear: 3-4                  8,320\n",
       "│    │    └─Linear: 3-5                  66,048\n",
       "│    │    └─Linear: 3-6                  402,192\n",
       "=================================================================\n",
       "Total params: 952,400\n",
       "Trainable params: 952,400\n",
       "Non-trainable params: 0\n",
       "================================================================="
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autoencoder = Autoencoder([512, 128, 64]).to(device)\n",
    "autoencoder = torch.compile(autoencoder)\n",
    "optim = torch.optim.Adam(autoencoder.parameters())\n",
    "summary(autoencoder)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80c8160f",
   "metadata": {},
   "source": [
    "Same number of parameters as for the Flax implementation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7da0b8f5",
   "metadata": {},
   "source": [
    "### 2.1.2. Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32aa5b6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "@torch.compile\n",
    "def compute_loss(model,  batch):\n",
    "    x = batch[0].to(device)\n",
    "    x_prime = model(x)\n",
    "    loss = F.binary_cross_entropy_with_logits(x_prime, x)\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4da92f42",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9e07e326fea2466980cd2b0b4c614190",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2023-09-15 22:44:59,656] torch._inductor.utils: [WARNING] DeviceCopy in input program\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2min 30s, sys: 1min 8s, total: 3min 38s\n",
      "Wall time: 4min 2s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    step_loss = 0\n",
    "    for batch in trainloader:\n",
    "        optim.zero_grad()\n",
    "        loss = compute_loss(autoencoder,  batch)\n",
    "        step_loss += loss.item()\n",
    "        loss.backward()\n",
    "        optim.step()\n",
    "    train_loss.append(step_loss / len(trainloader))\n",
    "    \n",
    "    step_loss = 0\n",
    "    with torch.no_grad():\n",
    "        for batch in testloader:\n",
    "            step_loss += compute_loss(autoencoder,  batch).item()\n",
    "        test_loss.append(step_loss / len(testloader))\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "781fd286",
   "metadata": {},
   "source": [
    "### 2.1.3. Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b9441eea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(torch.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(torch.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Learning curves for the autoencoder (PyTorch)\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91a5ba69",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "rec = F.sigmoid(autoencoder(imgs.to(device))).cpu().detach().numpy()\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = imgs[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = rec[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4c80a3d",
   "metadata": {},
   "source": [
    "The results are quite similar to the ones obtained with Flax both in terms of loss values and reconstructed images (apart from the noticeable white dot observed on the second image reconstructed). However, the training time is quite higher with PyTorch, even after compiling the model and trying to jit the loss computation.\n",
    "\n",
    "<ins>Training wall time</ins> :\n",
    "\n",
    "|   | Autoencodeur | VAE |\n",
    "|:-:|:-:|:-:|\n",
    "| Flax | <font color='LimeGreen'><b>2min 13s</b></font> | 2min 33s |\n",
    "| PyTorch | 4min 2s |   |\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f86f961d",
   "metadata": {},
   "source": [
    "## 2.2. VAE"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bebd76c7",
   "metadata": {},
   "source": [
    "### 2.2.1. Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f7f4f14e",
   "metadata": {},
   "outputs": [],
   "source": [
    "class VAE(nn.Module):\n",
    "    def __init__(self, latents):\n",
    "        \"\"\" Autoencoder constructor \"\"\"\n",
    "        super(VAE, self).__init__()\n",
    "        self.latents = [784] + latents\n",
    "        self.encoder = nn.Sequential()\n",
    "        self.decoder = nn.Sequential()\n",
    "        for i, l in enumerate(self.latents[:-2]):\n",
    "            self.encoder.append(nn.Linear(l, self.latents[i+1]))\n",
    "            self.encoder.append(nn.ReLU())\n",
    "        for i, l in enumerate(self.latents[::-1][:-1]):\n",
    "            self.decoder.append(nn.Linear(l, self.latents[::-1][i+1]))\n",
    "            if i != len(self.latents) - 2:\n",
    "                self.decoder.append(nn.ReLU())\n",
    "        self.mean = nn.Linear(self.latents[-2], self.latents[-1])\n",
    "        self.logvar = nn.Linear(self.latents[-2], self.latents[-1])\n",
    "                   \n",
    "    def forward(self, x):\n",
    "        x = self.encoder(x)\n",
    "        mean = self.mean(x)\n",
    "        logvar = self.logvar(x)\n",
    "        z = self.sample(mean, logvar)\n",
    "        return self.decoder(z), mean, logvar\n",
    "    \n",
    "    def sample(self, mean, logvar):\n",
    "        \"\"\" Sample latent vector from mean and log variance \"\"\"\n",
    "        std = torch.exp(0.5*logvar)  # 0.5 for root square, exp to cancel log\n",
    "        epsilon = torch.randn_like(std)\n",
    "        return mean + std * epsilon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6253f9fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "vae = VAE([512, 128, 64]).to(device)\n",
    "vae = torch.compile(vae)\n",
    "optim = torch.optim.Adam(vae.parameters())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d395c7b",
   "metadata": {},
   "source": [
    "### 2.2.2. Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2b5259a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "@torch.compile\n",
    "def compute_loss(model,  batch):\n",
    "    x = batch[0].to(device)\n",
    "    x_prime, mean, logvar = model(x)\n",
    "    cross_entropy = F.binary_cross_entropy_with_logits(x_prime, x, reduction='sum')\n",
    "    kl = -0.5 * torch.sum(1 + logvar - mean.pow(2) - logvar.exp())\n",
    "    return cross_entropy + kl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c9873bb2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9814d5a81367416296f27483a47b52de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2023-09-15 23:42:23,518] torch._inductor.utils: [WARNING] DeviceCopy in input program\n",
      "[2023-09-15 23:42:23,847] torch._inductor.utils: [WARNING] using triton random, expect difference from eager\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2min 37s, sys: 1min 4s, total: 3min 42s\n",
      "Wall time: 4min\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "epoch = tqdm(range(epochs))\n",
    "train_loss, test_loss = [], []\n",
    "for e in epoch:\n",
    "    torch.manual_seed(seed)\n",
    "    step_loss = 0\n",
    "    for batch in trainloader:\n",
    "        optim.zero_grad()\n",
    "        loss = compute_loss(vae,  batch)\n",
    "        step_loss += loss.item()\n",
    "        loss.backward()\n",
    "        optim.step()\n",
    "    train_loss.append(step_loss / len(trainloader))\n",
    "    \n",
    "    step_loss = 0\n",
    "    with torch.no_grad():\n",
    "        for batch in testloader:\n",
    "            step_loss += compute_loss(vae,  batch).item()\n",
    "        test_loss.append(step_loss / len(testloader))\n",
    "    epoch.set_description(f\"Epoch: {e+1}/{epochs} - Train Loss: {train_loss[-1]:.4f} - Test loss: {test_loss[-1]:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d806920",
   "metadata": {},
   "source": [
    "### 2.2.3 Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "74faeba3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(torch.arange(epochs), train_loss, label=\"Train loss\")\n",
    "plt.plot(torch.arange(epochs), test_loss, label=\"Test loss\")\n",
    "plt.title(\"Learning curves for the VAE (PyTorch)\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa366794",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imgs, _ = next(iter(testloader))\n",
    "rec, _, _ = vae(imgs.to(device))\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(10):\n",
    "    ax = plt.subplot(2, 10, i + 1)\n",
    "    img = imgs[i].reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Original\", fontweight=\"bold\")\n",
    "    ax = plt.subplot(2, 10, i + 11)\n",
    "    img = F.sigmoid(rec[i]).cpu().detach().numpy().reshape((28,28))\n",
    "    ax.imshow(img, cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(\"Reconstructed\", fontweight=\"bold\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2351f89",
   "metadata": {},
   "source": [
    "Once again, the images generated look quite similar to the ones obtained with Flax (the last `5` may be a bit worse). However, the learning curves seem to illustrate overfitting from around epoch 20 (even if the values obtained from the training loss are coherent with the ones obtained with Flax). <br>\n",
    "As for training time, Flax is definitely faster. The GPU used was a `NVIDIA GeForce RTX 3060 Laptop GPU` (6 Go)\n",
    "\n",
    "<ins>Training wall time</ins> :\n",
    "\n",
    "|   | Autoencodeur | VAE |\n",
    "|:-:|:-:|:-:|\n",
    "| Flax | <font color='LimeGreen'><b>2min 13s</b></font> | <font color='LimeGreen'><b>2min 33s</b></font> |\n",
    "| PyTorch | 4min 2s | 4min |"
   ]
  }
 ],
 "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": "3.11.2"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "428.8px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}