init

app.py

@@ -5,13 +5,13 @@ import torch
 from diffusers import AutoPipelineForText2Image, LCMScheduler
 from torchvision.transforms.functional import to_pil_image, center_crop, resize, to_tensor
 
-device = 'cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu'
+device = 'cpu'
 
 model_id = "Lykon/dreamshaper-7"
 adapter_id = "latent-consistency/lcm-lora-sdv1-5"
 
-# pipe = AutoPipelineForText2Image.from_pretrained(model_id, torch_dtype=torch.float16)
-pipe = AutoPipelineForText2Image.from_pretrained(model_id)
+pipe = AutoPipelineForText2Image.from_pretrained(model_id, torch_dtype=torch.float16)
+# pipe = AutoPipelineForText2Image.from_pretrained(model_id)
 pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
 pipe.to(device)
 
@@ -23,8 +23,8 @@ prompt = "Self-portrait oil painting, a beautiful cyborg with golden hair, 8k"
 
 
 @torch.no_grad()
-def generate(prompt, guidance_scale):
-    image = pipe(prompt=prompt, num_inference_steps=4, guidance_scale=guidance_scale).images[0]
+def generate(prompt, guidance_scale, num_inference_steps):
+    image = pipe(prompt=prompt, num_inference_steps=num_inference_steps, guidance_scale=guidance_scale).images[0]
     return image
 
 
@@ -37,7 +37,10 @@ def app():
                             value=prompt,
                         ),
                             gr.Slider(2, 20, value=7.5, label="Guidance Scale",
-                                      info="Higher scale depicts more creativity")],
+                                      info="Higher scale depicts more creativity"),
+                            gr.Slider(1, 50, value=4, label="Inference steps")
+                        ],
+
                         gr.Image(type="pil",
                                  height=512,
                                  width=512