fix cacheless case

src/transformers/generation/utils.py

@@ -693,8 +693,8 @@ def _update_model_kwargs_for_generation(
             model_kwargs["cache_position"] = model_kwargs["cache_position"][-1:] + num_new_tokens
         else:
             new_positions = torch.arange(
-                model_kwargs["cache_position"][-1],
-                model_kwargs["cache_position"][-1] + num_new_tokens,
+                model_kwargs["cache_position"][-1] + 1,
+                model_kwargs["cache_position"][-1] + num_new_tokens + 1,
                 device=model_kwargs["cache_position"].device,
                 dtype=model_kwargs["cache_position"].dtype,
             )

src/transformers/models/jamba/modeling_jamba.py

@@ -1553,7 +1553,7 @@ def prepare_inputs_for_generation(
         # If we have cache: let's slice `input_ids` through `cache_position`, to keep only the unprocessed tokens
         # Exception 1: when passing input_embeds, input_ids may be missing entries
         # Exception 2: some generation methods do special slicing of input_ids, so we don't need to do it here
-        if past_key_values is not None:
+        if not empty_past_kv:
             if inputs_embeds is not None:  # Exception 1
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)
@@ -1571,7 +1571,7 @@ def prepare_inputs_for_generation(
                 position_ids = position_ids[:, -input_ids.shape[1] :]
 
         # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
-        if inputs_embeds is not None and cache_position[0] == 0:
+        if inputs_embeds is not None and empty_past_kv:
             model_inputs = {"inputs_embeds": inputs_embeds}
         else:
             model_inputs = {"input_ids": input_ids.contiguous()}  # `contiguous()` needed for compilation use cases