Update generate.py

generate.py

@@ -2,10 +2,6 @@ import torch
 
 def generate(model, input_ids, generation_config, left_padding=None, **kwargs):
     generation_config = generation_config or model.generation_config  # default to the model generation config
-    pad_token = kwargs.pop("pad_token", None) or generation_config.pad_token_id or model.config.pad_token_id
-    if len(kwargs) > 0:  # Let's catch unexpected kwargs, so that users don't get surprised
-      raise ValueError(f"Unused kwargs: {list(kwargs.keys())}")
-
     cur_length = input_ids.shape[1]
     max_length = generation_config.max_length or cur_length + generation_config.max_new_tokens
 
@@ -14,6 +10,7 @@ def generate(model, input_ids, generation_config, left_padding=None, **kwargs):
         if not isinstance(left_padding, int) or left_padding < 0:
             raise ValueError(f"left_padding must be an integer larger than 0, but is {left_padding}")
 
+        pad_token = kwargs.pop("pad_token", None) or generation_config.pad_token_id or model.config.pad_token_id
         if pad_token is None:
             raise ValueError("pad_token is not defined")
         batch_size = input_ids.shape[0]