Update app.py

app.py

@@ -273,7 +273,7 @@ with gr.Blocks() as demo:
 
 # Initialize both pipelines
 if __name__ == "__main__":
-    from diffusers import FluxTransformer2DModel, FluxPipeline, BitsAndBytesConfig
+    from diffusers import FluxTransformer2DModel, FluxPipeline, BitsAndBytesConfig, GGUFQuantizationConfig
     from transformers import T5EncoderModel, BitsAndBytesConfig as BitsAndBytesConfigTF
     
     # Initialize Flux pipeline
@@ -282,11 +282,15 @@ if __name__ == "__main__":
     
     dtype = torch.bfloat16
     file_url = "https://huggingface.co/gokaygokay/flux-game/blob/main/gokaygokay_00001_.safetensors"
+    file_url = file_url.replace("/resolve/main/", "/blob/main/").replace("?download=true", "")
     single_file_base_model = "camenduru/FLUX.1-dev-diffusers"
     quantization_config_tf = BitsAndBytesConfigTF(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_use_double_quant=True, bnb_4bit_compute_dtype=torch.bfloat16)
     text_encoder_2 = T5EncoderModel.from_pretrained(single_file_base_model, subfolder="text_encoder_2", torch_dtype=dtype, config=single_file_base_model, quantization_config=quantization_config_tf, token=huggingface_token)
-    quantization_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_use_double_quant=True, bnb_4bit_compute_dtype=torch.bfloat16)
-    transformer = FluxTransformer2DModel.from_single_file(file_url, subfolder="transformer", torch_dtype=dtype, config=single_file_base_model, quantization_config=quantization_config, token=huggingface_token)
+    if ".gguf" in file_url:
+        transformer = transformer_model.from_single_file(file_url, subfolder="transformer", quantization_config=GGUFQuantizationConfig(compute_dtype=dtype), torch_dtype=dtype, config=single_file_base_model)
+    else:
+        quantization_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_use_double_quant=True, bnb_4bit_compute_dtype=torch.bfloat16, token=huggingface_token)
+        transformer = FluxTransformer2DModel.from_single_file(file_url, subfolder="transformer", torch_dtype=dtype, config=single_file_base_model, quantization_config=quantization_config, token=huggingface_token)
     flux_pipeline = FluxPipeline.from_pretrained(single_file_base_model, transformer=transformer, text_encoder_2=text_encoder_2, torch_dtype=dtype, quantization_config=quantization_config, token=huggingface_token)
 
     # Initialize Trellis pipeline