Update app.py

app.py

@@ -16,19 +16,13 @@ dtype = torch.bfloat16
 
 huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
 
-import torch
-from optimum.quanto import QuantizedDiffusersModel
-
-from diffusers.models.transformers.transformer_flux import FluxTransformer2DModel
-from diffusers.pipelines.flux.pipeline_flux import FluxPipeline
-
+from diffusers import FluxPipeline, FluxTransformer2DModel
 
-class QuantizedFluxTransformer2DModel(QuantizedDiffusersModel):
-    base_class = FluxTransformer2DModel
 
-
-transformer = QuantizedFluxTransformer2DModel.from_pretrained("Kijai/flux-fp8")
-transformer.to(device="cuda", dtype=torch.bfloat16)
+transformer = FluxTransformer2DModel.from_single_file(
+    "https://huggingface.co/Kijai/flux-fp8/blob/main/flux1-dev-fp8.safetensors",
+    torch_dtype=torch.float16,
+)
 
 pipe = FluxPipeline.from_pretrained(
     "black-forest-labs/FLUX.1-dev",