Update app.py

app.py

@@ -1,43 +1,108 @@
+# import gradio as gr
+# import torch
+# from diffusers import DiffusionPipeline, AutoPipelineForText2Image
+# import base64
+# from io import BytesIO
+
+
+
+# def load_amused_model():
+#     # pipeline = DiffusionPipeline.from_pretrained("Bakanayatsu/ponyDiffusion-V6-XL-Turbo-DPO") 
+#     # AutoPipelineForText2Image.from_pretrained("stabilityai/sd-turbo")
+#     # AutoPipelineForText2Image.from_pretrained("stabilityai/sdxl-turbo")
+#     return DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4",
+#         safety_checker = None,
+#         requires_safety_checker = False)
+     
+
+# # Generate image from prompt using AmusedPipeline
+# def generate_image(prompt):
+#     try:
+#         pipe = load_amused_model()
+#         generator = torch.Generator().manual_seed(8)  # Create a generator for reproducibility
+#         image = pipe(prompt, generator=generator).images[0]  # Generate image from prompt
+#         # image = pipe(prompt=prompt, num_inference_steps=1, guidance_scale=0.0).images[0]
+#         return image, None
+#     except Exception as e:
+#         return None, str(e)
+
+# def inference(prompt):
+#     print(f"Received prompt: {prompt}")  # Debugging statement
+#     image, error = generate_image(prompt)
+#     if error:
+#         print(f"Error generating image: {error}")  # Debugging statement
+#         return "Error: " + error
+    
+#     buffered = BytesIO()
+#     image.save(buffered, format="PNG")
+#     img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
+#     return img_str
+
+# gradio_interface = gr.Interface(
+#     fn=inference,
+#     inputs="text",
+#     outputs="text"  # Change output to text to return base64 string
+# )
+
+# if __name__ == "__main__":
+#     gradio_interface.launch()
+
+
+
 import gradio as gr
+from diffusers import DiffusionPipeline, DPMSolverSinglestepScheduler
 import torch
-from diffusers import DiffusionPipeline, AutoPipelineForText2Image
 import base64
 from io import BytesIO
 
 
-
 def load_amused_model():
     # pipeline = DiffusionPipeline.from_pretrained("Bakanayatsu/ponyDiffusion-V6-XL-Turbo-DPO") 
     # AutoPipelineForText2Image.from_pretrained("stabilityai/sd-turbo")
     # AutoPipelineForText2Image.from_pretrained("stabilityai/sdxl-turbo")
-    return DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4",
-        safety_checker = None,
-        requires_safety_checker = False)
-     
+    return DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float32).to("cpu")
+
 
 # Generate image from prompt using AmusedPipeline
 def generate_image(prompt):
     try:
         pipe = load_amused_model()
-        generator = torch.Generator().manual_seed(8)  # Create a generator for reproducibility
-        image = pipe(prompt, generator=generator).images[0]  # Generate image from prompt
+        pipe.load_lora_weights(
+            "mann-e/Mann-E_Turbo",
+            weight_name="manne_turbo.safetensors",
+        )
+        # This is equivalent to DPM++ SDE Karras, as noted in https://huggingface.co/docs/diffusers/main/en/api/schedulers/overview
+        pipe.scheduler = DPMSolverSinglestepScheduler.from_config(pipe.scheduler.config, use_karras_sigmas=True)
+
+        #generator = torch.Generator().manual_seed(8)  # Create a generator for reproducibility
+        #image = pipe(prompt, generator=generator).images[0]  # Generate image from prompt
         # image = pipe(prompt=prompt, num_inference_steps=1, guidance_scale=0.0).images[0]
+        image = pipe(
+            prompt="a cat in a bustling middle eastern city",
+            num_inference_steps=8,
+            guidance_scale=4,
+            width=768,
+            height=768,
+            clip_skip=1
+        ).images[0]
         return image, None
     except Exception as e:
         return None, str(e)
 
+
 def inference(prompt):
     print(f"Received prompt: {prompt}")  # Debugging statement
     image, error = generate_image(prompt)
     if error:
         print(f"Error generating image: {error}")  # Debugging statement
         return "Error: " + error
-    
+
     buffered = BytesIO()
     image.save(buffered, format="PNG")
     img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
     return img_str
 
+
 gradio_interface = gr.Interface(
     fn=inference,
     inputs="text",