Update app.py

app.py

@@ -36,11 +36,16 @@ sdxl_lightening = "ByteDance/SDXL-Lightning"
 ckpt = "sdxl_lightning_2step_unet.safetensors"
 unet = UNet2DConditionModel.from_config(model_id, subfolder="unet").to("cuda", torch.float16)
 unet.load_state_dict(load_file(hf_hub_download(sdxl_lightening, ckpt), device="cuda"))
-pipe = StableDiffusionXLPipeline.from_pretrained(model_id, unet=unet, torch_dtype=torch.float16, variant="fp16").to("cuda")
+pipe = SDEmb.from_pretrained(model_id, unet=unet, torch_dtype=torch.float16, variant="fp16").to("cuda")
 pipe.vae = AutoencoderTiny.from_pretrained("madebyollin/taesdxl", torch_dtype=torch.float16)
 pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing")
 pipe.to(device='cuda')
-pipe.load_ip_adapter("h94/IP-Adapter", subfolder="sdxl_models", weight_name="ip-adapter_sdxl.bin")
+
+image_encoder = CLIPimg.from_pretrained('h94/IP-Adapter', subfolder='sdxl_models/image_encoder', torch_dtype=torch.float16).to(device)
+pipe.unet._load_ip_adapter_weights(torch.load(hf_hub_download('h94/IP-Adapter', 'sdxl_models/ip-adapter_sdxl.bin'), map_location="cpu"))
+pipe.register_modules(image_encoder = image_encoder)
+
+# pipe.load_ip_adapter("h94/IP-Adapter", subfolder="sdxl_models", weight_name="ip-adapter_sdxl.bin")
 
 output_hidden_state = False
 #######################
@@ -57,7 +62,7 @@ def predict(
             im_emb = torch.zeros(1, 1280, dtype=torch.float16, device='cuda')
         image = pipe(
             prompt=prompt,
-            image_embeds=[im_emb.to('cuda')],
+            ip_adapter_emb=[im_emb.to('cuda')],
             height=1024,
             width=1024,                                                                                                             
             num_inference_steps=2,