Update app.py

app.py

@@ -2,12 +2,9 @@ from diffusers import DDPMPipeline
 image_pipe = DDPMPipeline.from_pretrained("google/ddpm-celebahq-256")
 image_pipe.to("cuda")
 images = image_pipe().images
-image_pipe
 from diffusers import UNet2DModel
 repo_id = "google/ddpm-church-256"
 model = UNet2DModel.from_pretrained(repo_id)
-model
-model.config
 model_random = UNet2DModel(**model.config)
 model_random.save_pretrained("my_model")
 model_random = UNet2DModel.from_pretrained("my_model")
@@ -16,19 +13,14 @@ torch.manual_seed(0)
 noisy_sample = torch.randn(
     1, model.config.in_channels, model.config.sample_size, model.config.sample_size
 )
-noisy_sample.shape
 with torch.no_grad():
     noisy_residual = model(sample=noisy_sample, timestep=2).sample
-noisy_residual.shape
 from diffusers import DDPMScheduler
 scheduler = DDPMScheduler.from_config(repo_id)
-scheduler.config
-scheduler.save_config("my_scheduler")
 new_scheduler = DDPMScheduler.from_config("my_scheduler")
 less_noisy_sample = scheduler.step(
     model_output=noisy_residual, timestep=2, sample=noisy_sample
 ).prev_sample
-less_noisy_sample.shape
 import PIL.Image
 import numpy as np
 def display_sample(sample, i):
@@ -38,7 +30,6 @@ def display_sample(sample, i):
     image_pil = PIL.Image.fromarray(image_processed[0])
     display(f"Image at step {i}")
     display(image_pil)
-model.to("cuda")
 noisy_sample = noisy_sample.to("cuda")
 import tqdm
 sample = noisy_sample
@@ -50,7 +41,6 @@ for i, t in enumerate(tqdm.tqdm(scheduler.timesteps)):
       display_sample(sample, i + 1)
 from diffusers import DDIMScheduler
 scheduler = DDIMScheduler.from_config(repo_id)
-scheduler.set_timesteps(num_inference_steps=50)
 import tqdm
 sample = noisy_sample
 for i, t in enumerate(tqdm.tqdm(scheduler.timesteps)):