Update punet to 4d4f955

Author: IanNod

models/turbine_models/custom_models/sdxl_inference/unet.py

@@ -92,7 +92,7 @@ def get_punet_model(hf_model_name, external_weight_path, precision="i8"):
     if precision == "i8":
         repo_id = "amd-shark/sdxl-quant-models"
         subfolder = "unet/int8"
-        revision = "d30d6ff79abb584bf2addc7866738df5242f315a"
+        revision = "4d4f95554bb991b95eb8ae4d57b38eb139b3e23f"
     elif precision in ["fp16", "fp32"]:
         repo_id = hf_model_name
         subfolder = "unet"