fix typo re model_path

a typo regarding model_path in `tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)`

README.md

@@ -55,7 +55,7 @@ model = AutoModel.from_pretrained(
     torch_dtype=torch.bfloat16,
     low_cpu_mem_usage=True,
     trust_remote_code=True).eval().cuda()
-tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)
+tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True, use_fast=False)
 generation_config = dict(max_new_tokens=1024, do_sample=True)