def main(args):
device = "cuda" #if args.use_cuda else "cpu"
# MODEL_DICT[args.model].reproduce(
# args.n_epochs, args.batch_size, args.log_dir, device
# )
train_loader, test_loader = datasets.get_mnist_loaders(
args.batch_size, dynamically_binarize=True)
model = models.PixelCNN(
in_channels=1,
out_channels=1,
n_residual=15,
residual_channels=16,
head_channels=32,
)
model.load_state_dict(torch.load(args.ckpt))
model = model.to(device)
sample_fn = model.sample
tensor = sample_fn(model)
if not os.path.exists(args.save_folder):
os.mkdir(args.save_folder)
tensor = torch.squeeze(tensor)
for i in range(len(tensor)):
img = tensor[i].detach().cpu().numpy()
#breakpoint()
cv2.imwrite(args.save_folder + '/' + str(i) + '.png', img * 255)
print(str(i), ' saved')
def test_PixelCNN(self):
model = models.PixelCNN(in_channels=3,
out_channels=3,
n_residual=1,
residual_channels=1,
head_channels=1)
self._smoke_test(model, in_channels=3)
def reproduce(n_epochs=457,
batch_size=256,
log_dir="/tmp/run",
device="cuda",
debug_loader=None):
"""Training script with defaults to reproduce results.
The code inside this function is self contained and can be used as a top level
training script, e.g. by copy/pasting it into a Jupyter notebook.
Args:
n_epochs: Number of epochs to train for.
batch_size: Batch size to use for training and evaluation.
log_dir: Directory where to log trainer state and TensorBoard summaries.
device: Device to train on (either 'cuda' or 'cpu').
debug_loader: Debug DataLoader which replaces the default training and
evaluation loaders if not 'None'. Do not use unless you're writing unit
tests.
"""
from torch import optim
from torch.nn import functional as F
from torch.optim import lr_scheduler
from pytorch_generative import datasets
from pytorch_generative import models
from pytorch_generative import trainer
train_loader, test_loader = debug_loader, debug_loader
if train_loader is None:
train_loader, test_loader = datasets.get_mnist_loaders(
batch_size, dynamically_binarize=True)
model = models.PixelCNN(
in_channels=1,
out_channels=1,
n_residual=15,
residual_channels=16,
head_channels=32,
)
optimizer = optim.Adam(model.parameters(), lr=1e-3)
scheduler = lr_scheduler.MultiplicativeLR(optimizer,
lr_lambda=lambda _: 0.999977)
def loss_fn(x, _, preds):
batch_size = x.shape[0]
x, preds = x.view((batch_size, -1)), preds.view((batch_size, -1))
loss = F.binary_cross_entropy_with_logits(preds, x, reduction="none")
return loss.sum(dim=1).mean()
trainer = trainer.Trainer(
model=model,
loss_fn=loss_fn,
optimizer=optimizer,
train_loader=train_loader,
eval_loader=test_loader,
lr_scheduler=scheduler,
log_dir=log_dir,
device=device,
)
trainer.interleaved_train_and_eval(n_epochs)
def test_PixelCNN_multiple_channels(self):
model = models.PixelCNN(
in_channels=3,
out_channels=3,
n_residual=1,
residual_channels=1,
head_channels=1,
)
self._smoke_test(model)
def test_PixelCNN(self):
model = models.PixelCNN(
in_channels=3,
out_channels=3,
n_residual=1,
residual_channels=1,
head_channels=1,
)
self._test_multiple_channels(model, conditional_sample=True)
def reproduce(
n_epochs=457, batch_size=256, log_dir="/tmp/run", device="cuda", debug_loader=None
):
"""Training script with defaults to reproduce results.
The code inside this function is self contained and can be used as a top level
training script, e.g. by copy/pasting it into a Jupyter notebook.
Args:
n_epochs: Number of epochs to train for.
batch_size: Batch size to use for training and evaluation.
log_dir: Directory where to log trainer state and TensorBoard summaries.
device: Device to train on (either 'cuda' or 'cpu').
debug_loader: Debug DataLoader which replaces the default training and
evaluation loaders if not 'None'. Do not use unless you're writing unit
tests.
"""
from torch import optim
from torch.nn import functional as F
from torch.optim import lr_scheduler
from torch.utils import data
from torchvision import datasets
from torchvision import transforms
from pytorch_generative import trainer
from pytorch_generative import models
####################################################################################################################
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~EB~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# Load ImageGPT Data :
import gmpm
import ipdb; ipdb.set_trace()
train = gmpm.train
test = gmpm.test
####################################################################################################################
transform = transforms.Compose(
[transforms.ToTensor(), lambda x: distributions.Bernoulli(probs=x).sample()]
)
train_loader = debug_loader or data.DataLoader(
datasets.MNIST("/tmp/data", train=True, download=True, transform=transform),
batch_size=batch_size,
shuffle=True,
num_workers=8,
)
test_loader = debug_loader or data.DataLoader(
datasets.MNIST("/tmp/data", train=False, download=True, transform=transform),
batch_size=batch_size,
num_workers=8,
)
model = models.PixelCNN(
in_channels=1,
out_channels=1,
n_residual=15,
residual_channels=16,
head_channels=32,
)
optimizer = optim.Adam(model.parameters(), lr=1e-3)
scheduler = lr_scheduler.MultiplicativeLR(optimizer, lr_lambda=lambda _: 0.999977)
def loss_fn(x, _, preds):
batch_size = x.shape[0]
x, preds = x.view((batch_size, -1)), preds.view((batch_size, -1))
loss = F.binary_cross_entropy_with_logits(preds, x, reduction="none")
return loss.sum(dim=1).mean()
trainer = trainer.Trainer(
model=model,
loss_fn=loss_fn,
optimizer=optimizer,
train_loader=train_loader,
eval_loader=test_loader,
lr_scheduler=scheduler,
log_dir=log_dir,
device=device,
)
trainer.interleaved_train_and_eval(n_epochs)