Skip to content

Instantly share code, notes, and snippets.

@neftlon
Created March 18, 2024 16:12
Show Gist options
  • Save neftlon/c57b0793ec3d1e64c7a206721bb2abeb to your computer and use it in GitHub Desktop.
Save neftlon/c57b0793ec3d1e64c7a206721bb2abeb to your computer and use it in GitHub Desktop.
VAE for MNIST
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "467f7ff2-d4c8-4df7-91ea-4edcac38e507",
"metadata": {},
"outputs": [],
"source": [
"import jax, jax.numpy as jnp, jax.random as jr, typing, matplotlib.pyplot as plt, tqdm, pandas as pd, datasets, math\n",
"\n",
"key = jr.key(42)"
]
},
{
"cell_type": "markdown",
"id": "00fc1708-034f-4862-9921-e1beb99e0e63",
"metadata": {},
"source": [
"Introduction from [here](https://arxiv.org/pdf/1906.02691.pdf).\n",
"\n",
"* **Latent variables:** if there is a latent variable $z$, the model can be expressed using the marginal likelihood $$p_\\theta(x)=\\int p_\\theta(x,z)\\mathrm{d}z.$$ Note that the posterior distribution $p_\\theta(z\\vert x)=\\frac{p_\\theta(x,z)}{p_\\theta(z)}$ is usually intractable because $p_\\theta(x)$ is usually intractable. Therefore, we cannot differentiate it w.r.t. its parameteters.\n",
"* **Deep latent variable model:** a latent variable model $p_\\theta(x,z)$ whose distribution is parametrized by a NN.\n",
"* **Approximate posterior:** In DLVMs, the posterior is intractable. It is therefore optimized through a parametric inference model such that $$q_\\phi(z\\vert x)\\approx p_\\theta(z\\vert x),$$ where $\\phi$ denotes the variational parameters.\n",
"* **Evidence lower bound (ELBO):** The ELBO can be derived using: $$\\begin{aligned}\\log p_\\theta(x)&=\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log p_\\theta(x)\\right]\\\\&=\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log\\frac{p_\\theta(x,z)}{p_\\theta(z\\vert x)}\\right]\\\\&=\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log\\left(\\frac{p_\\theta(x,z)}{q_\\phi(z\\vert x)}\\cdot\\frac{q_\\phi(z\\vert x)}{p_\\theta(z\\vert x)}\\right)\\right]\\\\&=\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log\\frac{p_\\theta(x,z)}{q_\\phi(z\\vert x)}\\right]+\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log\\frac{q_\\phi(z\\vert x)}{p_\\theta(z\\vert x)}\\right]\\\\&=\\mathcal{L}_{\\theta,\\phi}(x)+\\mathsf{KL}(q_\\phi(z\\vert x)\\vert\\vert p_\\theta(z\\vert x)).\\end{aligned}$$ Due to $\\mathsf{KL}(q_\\phi(z\\vert x)\\vert\\vert p_\\theta(z\\vert x)\\ge 0,$ the ELBO $$\\mathcal{L}_{\\theta,\\phi}(x)=\\mathbb{E}_{q_\\phi(z\\vert x)}\\left[\\log p_\\theta(x,z)-\\log q_\\phi(z\\vert x)\\right]$$ is a lower bound of the data's log-likelihood ($\\mathcal{L}_{\\theta,\\phi}(x)=\\log p_\\theta(x)-\\mathsf{KL}(q_\\phi(z\\vert x))\\vert\\vert p_\\theta(z\\vert x))\\le\\log p_\\theta(x)$). The Kullback-Leibler distance determines the distance of the approximate posterior from the true posterior by defining the tightness of the bound.\n",
"* **Variational autoencoder (VAE):** For VAEs, the variational distribution does not depend on the initial sample, but it's parameters do since $\\theta=\\mathtt{decoder}_\\psi(z)$ and $\\phi=\\mathtt{encoder}_\\lambda(x)$: $$\\mathcal{L}_{\\phi,\\theta}(x)=\\mathbb{E}_{q_\\phi(z)}\\left[\\log p_\\theta(x\\vert z)\\right]-\\mathsf{KL}(q_\\phi(z)\\vert\\vert p(z)).$$ We cannot compute the gradient $\\nabla_\\phi\\mathcal{L}_{\\phi,\\theta}(x)$ since it requires differentiating the expectation of w.r.t. to its parameters. By assuming a Gaussian $q_\\phi(z),$ we can reparametrize the distribution and rewrite the expectation. With this, the ELBO takes the following form: $$\\mathcal{L}_{\\phi,\\theta}(x)=\\mathbb{E}_{\\mathcal{N}(\\varepsilon\\vert 0,I)}\\left[\\log p_\\theta(x\\vert\\sigma\\odot\\varepsilon+\\mu)\\right]-\\mathsf{KL}(q_\\phi(z)\\vert\\vert p(z)),$$ where $(\\mu,\\log\\sigma)=\\phi$ is the output of the encoder. For standard normal Gaussian $p(z)$ and diagonal covariance Gaussian $q_\\phi(z)$, the KL-divergence can be found [here](https://en.wikipedia.org/wiki/Kullback%E2%80%93Leibler_divergence#Multivariate_normal_distributions)."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "99543867-6ec8-4446-8995-6f008d822981",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(60000, 784)"
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# load dataset\n",
"dataset = datasets.load_dataset(\"mnist\", split=\"train\")\n",
"X = dataset[\"image\"] # extract images\n",
"# binarize and flatten images\n",
"X = jnp.stack([(jnp.asarray(x) > 128).astype(float).ravel() for x in X])\n",
"X.shape"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "3143831b-d005-4da8-940d-fa0ada5b1cea",
"metadata": {},
"outputs": [],
"source": [
"class StandardNormal(typing.NamedTuple):\n",
" dim: int\n",
"\n",
" def sample(self, params, noise=None, key=None, num_samples: int = None):\n",
" if noise is not None:\n",
" # sample with reparametrization\n",
" assert key is None and num_samples is None\n",
" elif key is not None:\n",
" assert noise is None\n",
" shape_prefix = () if num_samples is None else (num_samples,)\n",
" noise = jr.normal(key, shape_prefix + (self.dim,))\n",
" else:\n",
" raise ValueError(\"either `key` or `noise` must not be None\")\n",
" return noise\n",
"\n",
"class DiagCovNormal(typing.NamedTuple):\n",
" dim: int\n",
"\n",
" def logpdf(self, params, x):\n",
" # compute logpdf for vectors stored in last dimension\n",
" # params: (B, 2 * self.dim) containing a batch of parameters\n",
" mu, logsig = jnp.moveaxis(params.reshape(-1,self.dim,2), -1, 0) # mu, logsig: (B,self.dim)\n",
" x = (x - mu) * jnp.exp(-logsig) # transform x (B, self.dim)\n",
" return self.dim / 2 * math.log(2 * math.pi) - jnp.einsum(\"...i,...j->...\",x,x)\n",
"\n",
" def sample(self, params, noise=None, key=None, num_samples: int = None):\n",
" mu, logsig = jnp.moveaxis(params.reshape(-1,self.dim,2), -1, 0) # mu, logsig: (B,self.dim)\n",
" if noise is not None:\n",
" # sample with reparametrization\n",
" assert key is None and num_samples is None\n",
" elif key is not None:\n",
" assert noise is None\n",
" shape_prefix = () if num_samples is None else (num_samples,)\n",
" noise_shape = shape_prefix + (self.dim,)\n",
" print(shape_prefix, noise_shape)\n",
" noise = jr.normal(key, noise_shape)\n",
" else:\n",
" raise ValueError(\"either `key` or `noise` must not be None\")\n",
" # reparametrize distribution\n",
" return mu + noise * jnp.exp(logsig)\n",
"\n",
" def kl_to(self, params, other, other_params):\n",
" if isinstance(other, StandardNormal):\n",
" assert self.dim == other.dim\n",
" del other_params # we don't need these for standard normal (they are tuple() anyway)\n",
" mu, logsig = jnp.moveaxis(params.reshape(-1,self.dim,2), -1, 0) # these can be of shapes (..., self.dim)\n",
" return .5 * jnp.sum(jnp.exp(2 * logsig) + mu ** 2 - 1 - 2 * logsig, axis=-1)\n",
" else:\n",
" raise ValueError(\"KL divergence between DiagCovNormal and %s is not supported\" % other.__class__.__name__)\n",
"\n",
"class Bernoulli(typing.NamedTuple):\n",
" dim: int\n",
"\n",
" def logpdf(self, params, x):\n",
" # compute logpdf for vectors stored in last dimension\n",
" # params: (B,L) containing parameters\n",
" return jnp.sum(params * jnp.log(x) + (1 - params) * jnp.log(1 - x), axis=-1)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "705e607f-a932-4513-a817-d281a4513e61",
"metadata": {},
"outputs": [],
"source": [
"class Mlp(typing.NamedTuple):\n",
" sizes: typing.Sequence[int]\n",
" activation: typing.Callable = jax.nn.leaky_relu\n",
" outactivation: typing.Callable = None\n",
"\n",
" def __call__(self, params, x):\n",
" for w, b in params[:-1]:\n",
" x = jnp.dot(x, w) + b\n",
" x = self.activation(x)\n",
" w, b = params[-1]\n",
" x = jnp.dot(x, w) + b\n",
" return self.outactivation(x) if self.outactivation is not None else x\n",
" \n",
" def init(self, key):\n",
" keys = jr.split(key, (len(self.sizes) - 1, 2))\n",
" z = zip(keys, self.sizes[:-1], self.sizes[1:])\n",
" return [(jr.normal(w, (m, n)) / jnp.sqrt(n * m), jr.normal(b, (n,)) / n) for (w, b), m, n in z]\n",
"\n",
" def __hash__(self):\n",
" tup = tuple(self.sizes), self.activation\n",
" return hash(tup)"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "a5fd3f33-92c8-4b48-9e1f-bc06e61c5f80",
"metadata": {},
"outputs": [],
"source": [
"class Vae(typing.NamedTuple):\n",
" encoder: typing.Callable\n",
" decoder: typing.Callable\n",
" target_distr: typing.Any # target distribution\n",
" latent_varia: typing.Any # latent variational distribution\n",
" latent_prior: typing.Any # prior distribution of the latent space\n",
" latent_prior_params: typing.Optional[typing.Any] = None\n",
"\n",
" def elbo(self, params, batch, eps, return_stats=False):\n",
" phi, theta = params # parameters of encoder and decoder\n",
"\n",
" # create latent samples\n",
" phi_params = self.encoder(phi, batch)\n",
" kls = self.latent_varia.kl_to(phi_params, self.latent_prior, self.latent_prior_params)\n",
" zs = self.latent_varia.sample(phi_params, noise=eps)\n",
"\n",
" # compute log likelihood\n",
" ys = self.decoder(theta, zs)\n",
" logliks = self.target_distr.logpdf(batch, ys)\n",
"\n",
" # average ELBO over batches\n",
" loglik, kl = jnp.mean(logliks), jnp.mean(kls)\n",
" res = loglik - kl,\n",
" if return_stats:\n",
" res = res + ((loglik, kl),)\n",
" return res\n",
"\n",
" def sample(self, params, key, num_samples):\n",
" _, theta = params\n",
" eps = self.latent_prior.sample(self.latent_prior_params, key=key, num_samples=num_samples)\n",
" return self.decoder(theta, eps)\n",
"\n",
" def init(self, key):\n",
" keys = jr.split(key)\n",
" return self.encoder.init(keys[0]), self.decoder.init(keys[1])\n",
"\n",
" @classmethod\n",
" def stdnorm_prior(cls, encoder, decoder, target_distr, latent_varia, latent_dim: int):\n",
" return cls(encoder, decoder, target_distr, latent_varia, StandardNormal(latent_dim), None)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "d7fd8c30-b4e0-48d2-9168-6cfe6ea8954c",
"metadata": {},
"outputs": [],
"source": [
"N, D = X.shape\n",
"L = 10 # latent space size\n",
"\n",
"# create and initialize model\n",
"vae = Vae.stdnorm_prior(\n",
" Mlp([D, 500, 2 * L]),\n",
" Mlp([L, 500, D], outactivation=jax.nn.sigmoid),\n",
" Bernoulli(D),\n",
" DiagCovNormal(L),\n",
" L,\n",
")\n",
"params_init = vae.init(key)\n",
"\n",
"# initialize training\n",
"params, elbos, logliks, kls = params_init, [], [], []"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "9b48e1c0-068c-4044-b7bb-dc1f6a69c216",
"metadata": {},
"outputs": [],
"source": [
"def aevb(key, model, params, xs, num_steps=1000, batch_size=100):\n",
" @jax.jit\n",
" def update_step(params, key, xs):\n",
" key, batch_key = jr.split(key)\n",
" batch = xs[jr.randint(batch_key, (batch_size,), 0, len(xs))]\n",
"\n",
" eps = model.latent_prior.sample(model.latent_prior_params, key=key, num_samples=batch_size) # noise sample\n",
" values, grads = jax.value_and_grad(model.elbo, has_aux=True)(params, batch, eps, True) # evaluate ELBO and compute gradients\n",
" elbo_val, (loglik_val, kl_val) = values # extract tracked variables\n",
" params = jax.tree.map(lambda p, g: p + 1e-3 * g, params, grads) # gradient ascent to maximize likelihood\n",
" return params, elbo_val, loglik_val, kl_val\n",
"\n",
" tracked_values = [], [], []\n",
" for _ in tqdm.trange(num_steps):\n",
" key, update_key = jr.split(key)\n",
" params, *vals = update_step(params, update_key, xs)\n",
" for hist, val in zip(tracked_values, vals):\n",
" hist.append(val.item())\n",
" return params, *tracked_values"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "f735b7b6-7c02-4e0c-9167-c600e1413b1f",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████████████████████████████████████████████████████████████████████████| 1000/1000 [00:01<00:00, 972.23it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-102.8636474609375\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 1000x500 with 3 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# run training\n",
"key, train_key = jr.split(key)\n",
"params, batch_elbos, batch_logliks, batch_kls = aevb(train_key, vae, params, X, num_steps=1000)\n",
"elbos += batch_elbos\n",
"logliks += batch_logliks\n",
"kls += batch_kls\n",
"\n",
"# print and plot history\n",
"print(elbos[-1])\n",
"plt.figure(figsize=(10,5))\n",
"for idx, (vals, names) in enumerate((\n",
" (elbos, r\"ELBO $\\mathcal{L}(\\phi,\\theta)$\"),\n",
" (logliks, r\"$\\log p(x\\vert z)$\"),\n",
" (kls, r\"$\\mathsf{KL}(q_\\phi(z)\\vert\\vert p(z))$\"),\n",
")):\n",
" plt.subplot(3,1,1 + idx)\n",
" plt.axhline(0, ls=\"--\", c=\"k\")\n",
" pd.Series(vals).plot.line(title=names)\n",
"plt.tight_layout()"
]
},
{
"cell_type": "code",
"execution_count": 31,
"id": "a3f16e38-04be-4eb4-8f07-d1bd25fbbe1e",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 800x800 with 100 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"key, noise_key = jr.split(key)\n",
"per_row = 10\n",
"samples = vae.sample(params, noise_key, per_row * per_row).reshape(-1,28,28)\n",
"\n",
"fig, axs = plt.subplots(figsize=(8,8), nrows=per_row, ncols=per_row)\n",
"for ax, sample in zip(axs.flat, samples):\n",
" plt.axis(\"off\")\n",
" ax.set(xticks=[], yticks=[])\n",
" ax.imshow(sample,cmap=\"gray\",interpolation=\"nearest\",vmin=0,vmax=1)\n",
"plt.tight_layout()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "1ab34b5a-a0c1-4ae6-b861-3b1a9b4b4bb9",
"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": "3.12.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment