Update app.py

Retesting Cascade

app.py

@@ -3,28 +3,36 @@ import torch
 import numpy as np
 import modin.pandas as pd
 from PIL import Image
-from diffusers import DiffusionPipeline
+#from diffusers import DiffusionPipeline
 from huggingface_hub import hf_hub_download
 
+from diffusers import StableCascadeCombinedPipeline
+
+
 device = 'cuda' #if torch.cuda.is_available() else 'cpu'
 torch.cuda.max_memory_allocated(device=device)
 torch.cuda.empty_cache()
-pipe = DiffusionPipeline.from_pretrained("circulus/canvers-fusionXL-v1", torch_dtype=torch.float16, safety_checker=None)
+pipe = StableCascadeCombinedPipeline.from_pretrained("stabilityai/stable-cascade", variant="bf16", torch_dtype=torch.bfloat16)
+
 pipe.enable_xformers_memory_efficient_attention()
 pipe = pipe.to(device)
 torch.cuda.empty_cache()
             
-refiner = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-refiner-1.0", use_safetensors=True, torch_dtype=torch.float16, variant="fp16") if torch.cuda.is_available() else DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-refiner-1.0")
-refiner.enable_xformers_memory_efficient_attention()
-refiner = refiner.to(device)
-torch.cuda.empty_cache()
-            
+         
 def genie (Prompt, negative_prompt, height, width, scale, steps, seed, upscale):
     generator = np.random.seed(0) if seed == 0 else torch.manual_seed(seed)
     torch.cuda.empty_cache()
-    int_image = pipe(prompt=Prompt, negative_prompt=negative_prompt, num_inference_steps=steps, guidance_scale=scale, width=width, height=height, output_type="latent").images #
-    torch.cuda.empty_cache()
-    image = refiner(Prompt, negative_prompt=negative_prompt, image=int_image, denoising_start=.99).images[0]
+    image=pipe(prompt=Prompt,
+    negative_prompt="",
+    num_inference_steps=20,
+    prior_num_inference_steps=steps,
+    prior_guidance_scale=scale,
+    width=width,
+    height=height,).images[0]
+    
+    #int_image = pipe(prompt=Prompt, negative_prompt=negative_prompt, num_inference_steps=steps, guidance_scale=scale, width=width, height=height, output_type="latent").images #
+    #torch.cuda.empty_cache()
+    #image = refiner(Prompt, negative_prompt=negative_prompt, image=int_image, denoising_start=.99).images[0]
     torch.cuda.empty_cache()
     return image
     
@@ -32,7 +40,7 @@ gr.Interface(fn=genie, inputs=[gr.Textbox(label='What you want the AI to generat
                                gr.Textbox(label='What you Do Not want the AI to generate. 77 Token Limit'),
                                gr.Slider(512, 1408, 1024, step=128, label='Height'),
                                gr.Slider(512, 1408, 1024, step=128, label='Width'),
-                               gr.Slider(7.5, maximum=15, value=10, step=.25, label='Guidance Scale'), 
+                               gr.Slider(.5, maximum=15, value=3, step=.25, label='Guidance Scale'), 
                                gr.Slider(10, maximum=50, value=25, step=5, label='Number of Iterations'), 
                                gr.Slider(minimum=0, step=1, maximum=9999999999999999, randomize=True, label='Seed: 0 is Random')],
              outputs=gr.Image(label='Generated Image'),