Update modeling_quiet.py

modeling_quiet.py

@@ -987,13 +987,15 @@ class QuietModel(QuietPreTrainedModel):
             batch_size, seq_length = input_ids.shape
         elif inputs_embeds is not None:
             if isinstance(inputs_embeds, list):
-                batch_size, seq_length, _ = inputs_embeds[0].shape
+                batch_size, seq_length = inputs_embeds[0].shape
+                inputs_embeds = torch.stack(inputs_embeds, dim=0)
             else:
                 batch_size, seq_length, _ = inputs_embeds.shape
         else:
             raise ValueError("You have to specify either decoder_input_ids or decoder_inputs_embeds")
 
 
+
         if self.gradient_checkpointing and self.training:
             if use_cache:
                 logger.warning_once(