Update app.py

app.py

@@ -74,11 +74,12 @@ pyx = cyper.inline(code, fast_indexing=True, directives=dict(boundscheck=False,
 
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
     #vae=AutoencoderKL.from_pretrained("ford442/sdxl-vae-bf16", use_safetensors=True, subfolder='vae',token=True)
-vaeX=AutoencoderKL.from_pretrained("ford442/stable-diffusion-3.5-large-fp32", safety_checker=None, use_safetensors=True, subfolder='vae', low_cpu_mem_usage=False, torch_dtype=torch.float32, token=True)
+#vaeX=AutoencoderKL.from_pretrained("ford442/stable-diffusion-3.5-large-fp32", safety_checker=None, use_safetensors=True, subfolder='vae', low_cpu_mem_usage=False, torch_dtype=torch.float32, token=True)
 pipe = StableDiffusion3Pipeline.from_pretrained(
     #"stabilityai  #  stable-diffusion-3.5-large",
-    "ford442/stable-diffusion-3.5-large-bf16",
-    vae=None,
+    #"ford442/stable-diffusion-3.5-large-bf16",
+    "ford442/stable-diffusion-3.5-large-fp32",
+    #vae=None,
     #vae=AutoencoderKL.from_pretrained("ford442/stable-diffusion-3.5-large-fp32", use_safetensors=True, subfolder='vae',token=True),
      #scheduler = FlowMatchHeunDiscreteScheduler.from_pretrained('ford442/stable-diffusion-3.5-large-bf16', subfolder='scheduler',token=True),
     text_encoder=None, #CLIPTextModelWithProjection.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder', token=True),
@@ -89,19 +90,20 @@ pipe = StableDiffusion3Pipeline.from_pretrained(
   #  text_encoder_3=T5EncoderModel.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder_3',token=True),
     #tokenizer=CLIPTokenizer.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", add_prefix_space=True, subfolder="tokenizer", token=True),
     #tokenizer_2=CLIPTokenizer.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", add_prefix_space=True, subfolder="tokenizer_2", token=True),
+    #tokenizer_3=T5TokenizerFast.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", add_prefix_space=False, use_fast=True, subfolder="tokenizer_3", token=True),
     tokenizer_3=T5TokenizerFast.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", add_prefix_space=False, use_fast=True, subfolder="tokenizer_3", token=True),
     #torch_dtype=torch.bfloat16,
     #use_safetensors=False,
 )
-text_encoder=CLIPTextModelWithProjection.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder', token=True).to(torch.device("cuda:0"), dtype=torch.bfloat16)
-text_encoder_2=CLIPTextModelWithProjection.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder_2',token=True).to(torch.device("cuda:0"), dtype=torch.bfloat16)
-text_encoder_3=T5EncoderModel.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder_3',token=True).to(torch.device("cuda:0"), dtype=torch.bfloat16)
+text_encoder=CLIPTextModelWithProjection.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder', token=True).to(torch.device("cuda:0")) #, dtype=torch.bfloat16)
+text_encoder_2=CLIPTextModelWithProjection.from_pretrained("ford442/stable-diffusion-3.5-large-bf16", subfolder='text_encoder_2',token=True).to(torch.device("cuda:0")) #, dtype=torch.bfloat16)
+text_encoder_3=T5EncoderModel.from_pretrained("ford442/stable-diffusion-3.5-large-fp32", subfolder='text_encoder_3',token=True).to(torch.device("cuda:0")) #, dtype=torch.bfloat16)
     
 pipe.load_lora_weights("ford442/sdxl-vae-bf16", weight_name="LoRA/UltraReal.safetensors")
 
-pipe.to(device=device, dtype=torch.bfloat16)
+pipe.to(device=device) #, dtype=torch.bfloat16)
 #pipe.to(device)
-pipe.vae=vaeX.to('cpu')
+#pipe.vae=vaeX.to('cpu')
 upscaler_2 = UpscaleWithModel.from_pretrained("Kim2091/ClearRealityV1").to(torch.device('cpu'))
 
 MAX_SEED = np.iinfo(np.int32).max