Update README.md

README.md

@@ -46,9 +46,7 @@ def load_model(model_size, model_dir):
     llama_config = LlamaConfig(**config)
     model = LlamaForCausalLM(llama_config).to('cuda')
     tokenizer = AutoTokenizer.from_pretrained('meta-llama/Llama-2-7b-hf', padding_side="right", use_fast=False)
-
-    if exist_extra_para:
-        model = replace_with_learnable_binarylinear(model, scaling_pattern = "column", keep_parts = ["lm_head"])
+    model = replace_with_learnable_binarylinear(model, scaling_pattern = "column", keep_parts = ["lm_head"])
 
     weight_dict = {}
     ckpt_plist = [p for p in model_dir.iterdir() if p.suffix == '.bin']