How to generate images after training. Hope it could help you guys.

[CooperLuo32]

` from denoising_diffusion_pytorch import Unet, GaussianDiffusion, Trainer

model = Unet( dim = 64, dim_mults = (1, 2, 4, 8), flash_attn = True )

diffusion = GaussianDiffusion( model, image_size = 128, timesteps = 1000, # number of steps sampling_timesteps = 250 # number of sampling timesteps (using ddim for faster inference [see citation for ddim paper]) )

trainer = Trainer( diffusion, 'path/to/your/images', train_batch_size = 32, train_lr = 8e-5, train_num_steps = 700000, # total training steps gradient_accumulate_every = 2, # gradient accumulation steps ema_decay = 0.995, # exponential moving average decay amp = True, # turn on mixed precision calculate_fid = True # whether to calculate fid during training )

trainer.load(4) # load model-4.pt # load the checkpoint

sampled_images = diffusion.sample(batch_size=8)

samples_root = r"./samples" os.makedirs(samples_root , exist_ok=True) len_samples = len(os.listdir(samples_root))

for i in range(sampled_images.size(0)):

current_image_tensor = sampled_images [i]
current_image = Image.fromarray((current_image_tensor.cpu().permute(1, 2, 0).numpy() * 255).astype('uint8'))
file_name = f"output__image_{i + len_samples}.png"
current_image.save(os.path.join(os.getcwd(),"samples/" + file_name))

print("all samples are save in folder") `

that's the code I write for generating images after you got the best checkpoint.Hope it could help if it do help,leave your comment and let me know, thank you

[lchunleo]

@CooperLuo32 thanks for sharing. may i check for the input to the model , is it random noises as defined by "Image.fromarray((current_image_tensor.cpu().permute(1, 2, 0).numpy() * 255).astype('uint8'))"

also is it a need to have the Trainer class to load the checkpoints? As i don't wish to train further but to use the checkpoint to perform inference

for "sampled_images = diffusion.sample(batch_size=8)", may i know what does the sample refers to ?