import gradio as gr from base64 import b64encode import numpy import torch from diffusers import AutoencoderKL, LMSDiscreteScheduler, UNet2DConditionModel from PIL import Image from torch import autocast from torchvision import transforms as tfms from tqdm.auto import tqdm from transformers import CLIPTextModel, CLIPTokenizer, logging import torchvision.transforms as T torch.manual_seed(1) logging.set_verbosity_error() torch_device = "cpu" vae = AutoencoderKL.from_pretrained("CompVis/stable-diffusion-v1-4", subfolder="vae") # Load the tokenizer and text encoder to tokenize and encode the text. tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14") text_encoder = CLIPTextModel.from_pretrained("openai/clip-vit-large-patch14") # The UNet model for generating the latents. unet = UNet2DConditionModel.from_pretrained("CompVis/stable-diffusion-v1-4", subfolder="unet") # The noise scheduler scheduler = LMSDiscreteScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", num_train_timesteps=1000) vae = vae.to(torch_device) text_encoder = text_encoder.to(torch_device) unet = unet.to(torch_device); token_emb_layer = text_encoder.text_model.embeddings.token_embedding pos_emb_layer = text_encoder.text_model.embeddings.position_embedding position_ids = text_encoder.text_model.embeddings.position_ids[:, :77] position_embeddings = pos_emb_layer(position_ids) def pil_to_latent(input_im): # Single image -> single latent in a batch (so size 1, 4, 64, 64) with torch.no_grad(): latent = vae.encode(tfms.ToTensor()(input_im).unsqueeze(0).to(torch_device)*2-1) # Note scaling return 0.18215 * latent.latent_dist.sample() def latents_to_pil(latents): # bath of latents -> list of images latents = (1 / 0.18215) * latents with torch.no_grad(): image = vae.decode(latents).sample image = (image / 2 + 0.5).clamp(0, 1) image = image.detach().cpu().permute(0, 2, 3, 1).numpy() images = (image * 255).round().astype("uint8") pil_images = [Image.fromarray(image) for image in images] return pil_images def get_output_embeds(input_embeddings): # CLIP's text model uses causal mask, so we prepare it here: bsz, seq_len = input_embeddings.shape[:2] causal_attention_mask = text_encoder.text_model._build_causal_attention_mask(bsz, seq_len, dtype=input_embeddings.dtype) # Getting the output embeddings involves calling the model with passing output_hidden_states=True # so that it doesn't just return the pooled final predictions: encoder_outputs = text_encoder.text_model.encoder( inputs_embeds=input_embeddings, attention_mask=None, # We aren't using an attention mask so that can be None causal_attention_mask=causal_attention_mask.to(torch_device), output_attentions=None, output_hidden_states=True, # We want the output embs not the final output return_dict=None, ) # We're interested in the output hidden state only output = encoder_outputs[0] # There is a final layer norm we need to pass these through output = text_encoder.text_model.final_layer_norm(output) # And now they're ready! return output def generate_with_embs(text_embeddings, seed, max_length): height = 512 # default height of Stable Diffusion width = 512 # default width of Stable Diffusion num_inference_steps = 10 # Number of denoising steps guidance_scale = 7.5 # Scale for classifier-free guidance generator = torch.manual_seed(seed) # Seed generator to create the inital latent noise batch_size = 1 # max_length = text_input.input_ids.shape[-1] uncond_input = tokenizer( [""] * batch_size, padding="max_length", max_length=max_length, return_tensors="pt" ) with torch.no_grad(): uncond_embeddings = text_encoder(uncond_input.input_ids.to(torch_device))[0] text_embeddings = torch.cat([uncond_embeddings, text_embeddings]) # Prep Scheduler set_timesteps(scheduler, num_inference_steps) # Prep latents latents = torch.randn( (batch_size, unet.in_channels, height // 8, width // 8), generator=generator, ) latents = latents.to(torch_device) latents = latents * scheduler.init_noise_sigma # Loop for i, t in tqdm(enumerate(scheduler.timesteps), total=len(scheduler.timesteps)): # expand the latents if we are doing classifier-free guidance to avoid doing two forward passes. latent_model_input = torch.cat([latents] * 2) sigma = scheduler.sigmas[i] latent_model_input = scheduler.scale_model_input(latent_model_input, t) # predict the noise residual with torch.no_grad(): noise_pred = unet(latent_model_input, t, encoder_hidden_states=text_embeddings)["sample"] # perform guidance noise_pred_uncond, noise_pred_text = noise_pred.chunk(2) noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond) # compute the previous noisy sample x_t -> x_t-1 latents = scheduler.step(noise_pred, t, latents).prev_sample return latents_to_pil(latents)[0] # Prep Scheduler def set_timesteps(scheduler, num_inference_steps): scheduler.set_timesteps(num_inference_steps) scheduler.timesteps = scheduler.timesteps.to(torch.float32) # minor fix to ensure MPS compatibility, fixed in diffusers PR 3925 def embed_style(prompt, style_embed, style_seed): # Tokenize text_input = tokenizer(prompt, padding="max_length", max_length=tokenizer.model_max_length, truncation=True, return_tensors="pt") input_ids = text_input.input_ids.to(torch_device) # Get token embeddings token_embeddings = token_emb_layer(input_ids) replacement_token_embedding = style_embed.to(torch_device) # Insert this into the token embeddings token_embeddings[0, torch.where(input_ids[0]==6829)] = replacement_token_embedding.to(torch_device) # Combine with pos embs input_embeddings = token_embeddings + position_embeddings # Feed through to get final output embs modified_output_embeddings = get_output_embeds(input_embeddings) # And generate an image with this: max_length = text_input.input_ids.shape[-1] return generate_with_embs(modified_output_embeddings, style_seed, max_length) def loss_style(prompt, style_embed, style_seed): # Tokenize text_input = tokenizer(prompt, padding="max_length", max_length=tokenizer.model_max_length, truncation=True, return_tensors="pt") input_ids = text_input.input_ids.to(torch_device) # Get token embeddings token_embeddings = token_emb_layer(input_ids) # The new embedding - our special birb word replacement_token_embedding = style_embed.to(torch_device) # Insert this into the token embeddings token_embeddings[0, torch.where(input_ids[0]==6829)] = replacement_token_embedding.to(torch_device) # Combine with pos embs input_embeddings = token_embeddings + position_embeddings # Feed through to get final output embs modified_output_embeddings = get_output_embeds(input_embeddings) # And generate an image with this: max_length = text_input.input_ids.shape[-1] return generate_loss_based_image(modified_output_embeddings, style_seed,max_length) def color_loss(image): color_channel = image[:, 1] target_value = 0.7 error = torch.abs(color_channel - target_value).mean() return error def generate_loss_based_image(text_embeddings, seed, max_length): height = 64 width = 64 num_inference_steps = 10 guidance_scale = 8 generator = torch.manual_seed(64) batch_size = 1 loss_scale = 200 uncond_input = tokenizer( [""] * batch_size, padding="max_length", max_length=max_length, return_tensors="pt" ) with torch.no_grad(): uncond_embeddings = text_encoder(uncond_input.input_ids.to(torch_device))[0] text_embeddings = torch.cat([uncond_embeddings, text_embeddings]) # Prep Scheduler set_timesteps(scheduler, num_inference_steps+1) # Prep latents latents = torch.randn( (batch_size, unet.in_channels, height // 8, width // 8), generator=generator, ) latents = latents.to(torch_device) latents = latents * scheduler.init_noise_sigma sched_out = None # Loop for i, t in tqdm(enumerate(scheduler.timesteps), total=len(scheduler.timesteps)): # expand the latents if we are doing classifier-free guidance to avoid doing two forward passes. latent_model_input = torch.cat([latents] * 2) sigma = scheduler.sigmas[i] latent_model_input = scheduler.scale_model_input(latent_model_input, t) # predict the noise residual with torch.no_grad(): noise_pred = unet(latent_model_input, t, encoder_hidden_states=text_embeddings)["sample"] # perform CFG noise_pred_uncond, noise_pred_text = noise_pred.chunk(2) noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond) ### ADDITIONAL GUIDANCE ### if i%5 == 0 and i>0: # Requires grad on the latents latents = latents.detach().requires_grad_() # Get the predicted x0: scheduler._step_index -= 1 latents_x0 = scheduler.step(noise_pred, t, latents).pred_original_sample # Decode to image space denoised_images = vae.decode((1 / 0.18215) * latents_x0).sample / 2 + 0.5 # range (0, 1) # Calculate loss loss = color_loss(denoised_images) * loss_scale # Occasionally print it out # if i%10==0: print(i, 'loss:', loss) # Get gradient cond_grad = torch.autograd.grad(loss, latents)[0] # Modify the latents based on this gradient latents = latents.detach() - cond_grad * sigma**2 # To PIL Images im_t0 = latents_to_pil(latents_x0)[0] im_next = latents_to_pil(latents)[0] # Now step with scheduler latents = scheduler.step(noise_pred, t, latents).prev_sample return latents_to_pil(latents)[0] def generate_image_from_prompt(text_in, style_in): STYLE_LIST = ['coffeemachine.bin', 'collage_style.bin', 'cube.bin', 'jerrymouse2.bin', 'zero.bin'] STYLE_SEEDS = [32, 64, 128, 16, 8] print(text_in) print(style_in) style_file = style_in + '.bin' idx = STYLE_LIST.index(style_file) print(style_file) print(idx) prompt = text_in style_seed = STYLE_SEEDS[idx] style_dict = torch.load(style_file) style_embed = [v for v in style_dict.values()] generated_image = embed_style(prompt, style_embed[0], style_seed) loss_generated_img = (loss_style(prompt, style_embed[0], style_seed)) return [generated_image, loss_generated_img] # Define Interface title = 'ERA-SESSION20 Generative Art and Stable Diffusion' demo = gr.Interface(generate_image_from_prompt, inputs = [gr.Textbox(1, label='prompt'), gr.Dropdown( ['coffeemachine.bin', 'collage_style.bin', 'cube.bin', 'jerrymouse2.bin', 'zero.bin'],value="cube", label="Pretrained Styles" ) ], outputs = [ gr.Gallery(label="Generated images", show_label=False, elem_id="gallery", columns=[2], rows=[2], object_fit="contain", height="auto") ], title = title ) demo.launch(debug=True)