# Written by Francois Fleuret <francois@fleuret.org>
-# Minimal implementation of Jonathan Ho, Ajay Jain, Pieter Abbeel
-# "Denoising Diffusion Probabilistic Models" (2020)
-#
-# https://arxiv.org/abs/2006.11239
+import math, argparse
-import math
import matplotlib.pyplot as plt
-import torch
+
+import torch, torchvision
from torch import nn
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+print(f'device {device}')
+
+######################################################################
+
+def sample_gaussian_mixture(nb):
+ p, std = 0.3, 0.2
+ result = torch.randn(nb, 1) * std
+ result = result + torch.sign(torch.rand(result.size()) - p) / 2
+ return result
+
+def sample_ramp(nb):
+ result = torch.min(torch.rand(nb, 1), torch.rand(nb, 1))
+ return result
+
+def sample_two_discs(nb):
+ a = torch.rand(nb) * math.pi * 2
+ b = torch.rand(nb).sqrt()
+ q = (torch.rand(nb) <= 0.5).long()
+ b = b * (0.3 + 0.2 * q)
+ result = torch.empty(nb, 2)
+ result[:, 0] = a.cos() * b - 0.5 + q
+ result[:, 1] = a.sin() * b - 0.5 + q
+ return result
+
+def sample_disc_grid(nb):
+ a = torch.rand(nb) * math.pi * 2
+ b = torch.rand(nb).sqrt()
+ N = 4
+ q = (torch.randint(N, (nb,)) - (N - 1) / 2) / ((N - 1) / 2)
+ r = (torch.randint(N, (nb,)) - (N - 1) / 2) / ((N - 1) / 2)
+ b = b * 0.1
+ result = torch.empty(nb, 2)
+ result[:, 0] = a.cos() * b + q
+ result[:, 1] = a.sin() * b + r
+ return result
+
+def sample_spiral(nb):
+ u = torch.rand(nb)
+ rho = u * 0.65 + 0.25 + torch.rand(nb) * 0.15
+ theta = u * math.pi * 3
+ result = torch.empty(nb, 2)
+ result[:, 0] = theta.cos() * rho
+ result[:, 1] = theta.sin() * rho
+ return result
+
+def sample_mnist(nb):
+ train_set = torchvision.datasets.MNIST(root = './data/', train = True, download = True)
+ result = train_set.data[:nb].to(device).view(-1, 1, 28, 28).float()
+ return result
+
+samplers = {
+ 'gaussian_mixture': sample_gaussian_mixture,
+ 'ramp': sample_ramp,
+ 'two_discs': sample_two_discs,
+ 'disc_grid': sample_disc_grid,
+ 'spiral': sample_spiral,
+ 'mnist': sample_mnist,
+}
+
+######################################################################
+
+parser = argparse.ArgumentParser(
+ description = '''A minimal implementation of Jonathan Ho, Ajay Jain, Pieter Abbeel
+"Denoising Diffusion Probabilistic Models" (2020)
+https://arxiv.org/abs/2006.11239''',
+
+ formatter_class = argparse.ArgumentDefaultsHelpFormatter
+)
+
+parser.add_argument('--seed',
+ type = int, default = 0,
+ help = 'Random seed, < 0 is no seeding')
+
+parser.add_argument('--nb_epochs',
+ type = int, default = 100,
+ help = 'How many epochs')
+
+parser.add_argument('--batch_size',
+ type = int, default = 25,
+ help = 'Batch size')
+
+parser.add_argument('--nb_samples',
+ type = int, default = 25000,
+ help = 'Number of training examples')
+
+parser.add_argument('--learning_rate',
+ type = float, default = 1e-3,
+ help = 'Learning rate')
+
+parser.add_argument('--ema_decay',
+ type = float, default = 0.9999,
+ help = 'EMA decay, <= 0 is no EMA')
+
+data_list = ', '.join( [ str(k) for k in samplers ])
+
+parser.add_argument('--data',
+ type = str, default = 'gaussian_mixture',
+ help = f'Toy data-set to use: {data_list}')
+
+args = parser.parse_args()
+
+if args.seed >= 0:
+ # torch.backends.cudnn.deterministic = True
+ # torch.backends.cudnn.benchmark = False
+ # torch.use_deterministic_algorithms(True)
+ torch.manual_seed(args.seed)
+ if torch.cuda.is_available():
+ torch.cuda.manual_seed_all(args.seed)
+
######################################################################
class EMA:
- def __init__(self, model, decay = 0.9999):
+ def __init__(self, model, decay):
self.model = model
self.decay = decay
- self.ema = { }
+ self.mem = { }
with torch.no_grad():
for p in model.parameters():
- self.ema[p] = p.clone()
+ self.mem[p] = p.clone()
def step(self):
with torch.no_grad():
for p in self.model.parameters():
- self.ema[p].copy_(self.decay * self.ema[p] + (1 - self.decay) * p)
+ self.mem[p].copy_(self.decay * self.mem[p] + (1 - self.decay) * p)
- def copy(self):
+ def copy_to_model(self):
with torch.no_grad():
for p in self.model.parameters():
- p.copy_(self.ema[p])
+ p.copy_(self.mem[p])
######################################################################
-def sample_gaussian_mixture(nb):
- p, std = 0.3, 0.2
- result = torch.empty(nb, 1).normal_(0, std)
- result = result + torch.sign(torch.rand(result.size()) - p) / 2
- return result
+class ConvNet(nn.Module):
+ def __init__(self, in_channels, out_channels):
+ super().__init__()
+
+ ks, nc = 5, 64
+
+ self.core = nn.Sequential(
+ nn.Conv2d(in_channels, nc, ks, padding = ks//2),
+ nn.ReLU(),
+ nn.Conv2d(nc, nc, ks, padding = ks//2),
+ nn.ReLU(),
+ nn.Conv2d(nc, nc, ks, padding = ks//2),
+ nn.ReLU(),
+ nn.Conv2d(nc, nc, ks, padding = ks//2),
+ nn.ReLU(),
+ nn.Conv2d(nc, nc, ks, padding = ks//2),
+ nn.ReLU(),
+ nn.Conv2d(nc, out_channels, ks, padding = ks//2),
+ )
+
+ def forward(self, x):
+ return self.core(x)
-def sample_arc(nb):
- theta = torch.rand(nb) * math.pi
- rho = torch.rand(nb) * 0.1 + 0.7
- result = torch.empty(nb, 2)
- result[:, 0] = theta.cos() * rho
- result[:, 1] = theta.sin() * rho
- return result
+######################################################################
+# Data
+
+try:
+ train_input = samplers[args.data](args.nb_samples).to(device)
+except KeyError:
+ print(f'unknown data {args.data}')
+ exit(1)
+
+train_mean, train_std = train_input.mean(), train_input.std()
######################################################################
-# Train
+# Model
+
+if train_input.dim() == 2:
+ nh = 256
+
+ model = nn.Sequential(
+ nn.Linear(train_input.size(1) + 1, nh),
+ nn.ReLU(),
+ nn.Linear(nh, nh),
+ nn.ReLU(),
+ nn.Linear(nh, nh),
+ nn.ReLU(),
+ nn.Linear(nh, train_input.size(1)),
+ )
-nb_samples = 25000
+elif train_input.dim() == 4:
-train_input = sample_gaussian_mixture(nb_samples)
-#train_input = sample_arc(nb_samples)
+ model = ConvNet(train_input.size(1) + 1, train_input.size(1))
+
+model.to(device)
+
+print(f'nb_parameters {sum([ p.numel() for p in model.parameters() ])}')
######################################################################
+# Generate
-nh = 64
+def generate(size, alpha, alpha_bar, sigma, model):
+ with torch.no_grad():
+ x = torch.randn(size, device = device)
-model = nn.Sequential(
- nn.Linear(train_input.size(1) + 1, nh),
- nn.ReLU(),
- nn.Linear(nh, nh),
- nn.ReLU(),
- nn.Linear(nh, train_input.size(1)),
-)
+ for t in range(T-1, -1, -1):
+ z = torch.zeros_like(x) if t == 0 else torch.randn_like(x)
+ input = torch.cat((x, torch.full_like(x[:,:1], t / (T - 1) - 0.5)), 1)
+ x = 1/torch.sqrt(alpha[t]) \
+ * (x - (1-alpha[t]) / torch.sqrt(1-alpha_bar[t]) * model(input)) \
+ + sigma[t] * z
+
+ x = x * train_std + train_mean
-nb_epochs = 50
-batch_size = 25
+ return x
+
+######################################################################
+# Train
T = 1000
-beta = torch.linspace(1e-4, 0.02, T)
+beta = torch.linspace(1e-4, 0.02, T, device = device)
alpha = 1 - beta
alpha_bar = alpha.log().cumsum(0).exp()
sigma = beta.sqrt()
-ema = EMA(model)
+ema = EMA(model, decay = args.ema_decay) if args.ema_decay > 0 else None
-for k in range(nb_epochs):
+for k in range(args.nb_epochs):
acc_loss = 0
- optimizer = torch.optim.Adam(model.parameters(), lr = 1e-3)
-
- for x0 in train_input.split(batch_size):
- t = torch.randint(T, (x0.size(0), 1))
- eps = torch.randn(x0.size())
- input = alpha_bar[t].sqrt() * x0 + (1 - alpha_bar[t]).sqrt() * eps
- input = torch.cat((input, 2 * t / T - 1), 1)
- output = model(input)
- loss = (eps - output).pow(2).mean()
+ optimizer = torch.optim.Adam(model.parameters(), lr = args.learning_rate)
+
+ for x0 in train_input.split(args.batch_size):
+ x0 = (x0 - train_mean) / train_std
+ t = torch.randint(T, (x0.size(0),) + (1,) * (x0.dim() - 1), device = x0.device)
+ eps = torch.randn_like(x0)
+ xt = torch.sqrt(alpha_bar[t]) * x0 + torch.sqrt(1 - alpha_bar[t]) * eps
+ input = torch.cat((xt, t.expand_as(x0[:,:1]) / (T - 1) - 0.5), 1)
+ loss = (eps - model(input)).pow(2).mean()
+ acc_loss += loss.item() * x0.size(0)
+
optimizer.zero_grad()
loss.backward()
optimizer.step()
- acc_loss += loss.item()
-
- ema.step()
+ if ema is not None: ema.step()
- if k%10 == 0: print(k, loss.item())
+ print(f'{k} {acc_loss / train_input.size(0)}')
-ema.copy()
+if ema is not None: ema.copy_to_model()
######################################################################
-# Generate
+# Plot
-x = torch.randn(10000, train_input.size(1))
+model.eval()
-for t in range(T-1, -1, -1):
- z = torch.zeros(x.size()) if t == 0 else torch.randn(x.size())
- input = torch.cat((x, torch.ones(x.size(0), 1) * 2 * t / T - 1), 1)
- x = 1 / alpha[t].sqrt() * (x - (1 - alpha[t])/(1 - alpha_bar[t]).sqrt() * model(input)) \
- + sigma[t] * z
+if train_input.dim() == 2:
-######################################################################
-# Plot
+ fig = plt.figure()
+ ax = fig.add_subplot(1, 1, 1)
+
+ if train_input.size(1) == 1:
+
+ x = generate((10000, 1), alpha, alpha_bar, sigma, model)
+
+ ax.set_xlim(-1.25, 1.25)
+ ax.spines.right.set_visible(False)
+ ax.spines.top.set_visible(False)
-fig = plt.figure()
-ax = fig.add_subplot(1, 1, 1)
+ d = train_input.flatten().detach().to('cpu').numpy()
+ ax.hist(d, 25, (-1, 1),
+ density = True,
+ histtype = 'stepfilled', color = 'lightblue', label = 'Train')
-if train_input.size(1) == 1:
+ d = x.flatten().detach().to('cpu').numpy()
+ ax.hist(d, 25, (-1, 1),
+ density = True,
+ histtype = 'step', color = 'red', label = 'Synthesis')
- ax.set_xlim(-1.25, 1.25)
+ ax.legend(frameon = False, loc = 2)
- d = train_input.flatten().detach().numpy()
- ax.hist(d, 25, (-1, 1),
- density = True,
- histtype = 'stepfilled', color = 'lightblue', label = 'Train')
+ elif train_input.size(1) == 2:
- d = x.flatten().detach().numpy()
- ax.hist(d, 25, (-1, 1),
- density = True,
- histtype = 'step', color = 'red', label = 'Synthesis')
+ x = generate((1000, 2), alpha, alpha_bar, sigma, model)
- ax.legend(frameon = False, loc = 2)
+ ax.set_xlim(-1.5, 1.5)
+ ax.set_ylim(-1.5, 1.5)
+ ax.set(aspect = 1)
+ ax.spines.right.set_visible(False)
+ ax.spines.top.set_visible(False)
-elif train_input.size(1) == 2:
+ d = x.detach().to('cpu').numpy()
+ ax.scatter(d[:, 0], d[:, 1],
+ s = 2.0, color = 'red', label = 'Synthesis')
- ax.set_xlim(-1.25, 1.25)
- ax.set_ylim(-1.25, 1.25)
- ax.set(aspect = 1)
+ d = train_input[:x.size(0)].detach().to('cpu').numpy()
+ ax.scatter(d[:, 0], d[:, 1],
+ s = 2.0, color = 'gray', label = 'Train')
- d = train_input[:200].detach().numpy()
- ax.scatter(d[:, 0], d[:, 1],
- color = 'lightblue', label = 'Train')
+ ax.legend(frameon = False, loc = 2)
- d = x[:200].detach().numpy()
- ax.scatter(d[:, 0], d[:, 1],
- color = 'red', label = 'Synthesis')
+ filename = f'diffusion_{args.data}.pdf'
+ print(f'saving {filename}')
+ fig.savefig(filename, bbox_inches='tight')
- ax.legend(frameon = False, loc = 2)
+ if hasattr(plt.get_current_fig_manager(), 'window'):
+ plt.get_current_fig_manager().window.setGeometry(2, 2, 1024, 768)
+ plt.show()
-filename = 'diffusion.pdf'
-print(f'saving {filename}')
-fig.savefig(filename, bbox_inches='tight')
+elif train_input.dim() == 4:
-plt.get_current_fig_manager().window.setGeometry(2, 2, 1024, 768)
-plt.show()
+ x = generate((128,) + train_input.size()[1:], alpha, alpha_bar, sigma, model)
+ x = 1 - x.clamp(min = 0, max = 255) / 255
+ torchvision.utils.save_image(x, f'diffusion_{args.data}.png', nrow = 16, pad_value = 0.8)
######################################################################