higher default learning rate

README.md

@@ -64,7 +64,7 @@ trainer = Trainer(
     diffusion,
     'path/to/your/images',
     train_batch_size = 32,
-    train_lr = 2e-5,
+    train_lr = 1e-4,
     train_num_steps = 700000,         # total training steps
     gradient_accumulate_every = 2,    # gradient accumulation steps
     ema_decay = 0.995,                # exponential moving average decay

denoising_diffusion_pytorch/denoising_diffusion_pytorch.py

@@ -440,7 +440,7 @@ class GaussianDiffusion(nn.Module):
         for i in tqdm(reversed(range(0, self.num_timesteps)), desc='sampling loop time step', total=self.num_timesteps):
             img = self.p_sample(img, torch.full((b,), i, device=device, dtype=torch.long))
 
-        img = normalize_to_neg_one_to_one(img)
+        img = unnormalize_to_zero_to_one(img)
         return img
 
     @torch.no_grad()
@@ -542,7 +542,7 @@ class Trainer(object):
         ema_decay = 0.995,
         image_size = 128,
         train_batch_size = 32,
-        train_lr = 2e-5,
+        train_lr = 1e-4,
         train_num_steps = 100000,
         gradient_accumulate_every = 2,
         amp = False,

setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 setup(
   name = 'denoising-diffusion-pytorch',
   packages = find_packages(),
-  version = '0.15.4',
+  version = '0.15.7',
   license='MIT',
   description = 'Denoising Diffusion Probabilistic Models - Pytorch',
   author = 'Phil Wang',