From 5a63326bb9207e803f3882b009e67c0f3dc0ca64 Mon Sep 17 00:00:00 2001 From: DeepBeepMeep Date: Fri, 13 Jun 2025 00:05:07 +0200 Subject: [PATCH] fixed autoquantize transformer --- wgp.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/wgp.py b/wgp.py index f16f958..d9e6b43 100644 --- a/wgp.py +++ b/wgp.py @@ -2293,7 +2293,7 @@ def load_models(model_type): model_filename = get_model_filename(model_type=model_type, quantization= transformer_quantization, dtype_policy = transformer_dtype_policy) base_model_type = get_base_model_type(model_type) finetune_def = get_model_finetune_def(model_type) - quantizeTransformer = finetune_def !=None and finetune_def.get("auto_quantize", False) and not "quanto" in model_filename + quantizeTransformer = finetune_def !=None and transformer_quantization in ("int8", "fp8") and finetune_def.get("auto_quantize", False) and not "quanto" in model_filename model_family = get_model_family(model_type) perc_reserved_mem_max = args.perc_reserved_mem_max