Correct the new defaults (#34377)

* Correct the new defaults

* CIs

* add check

* Update utils.py

* Update utils.py

* Add the max_length in generate test checking shape without passing length

* style

* CIs

* fix fx CI issue

src/transformers/generation/utils.py

@@ -1440,8 +1440,11 @@ class GenerationMixin:
             and not self.config.is_encoder_decoder
         ):
             generation_config.max_length -= inputs_tensor.shape[1]
-        else:  # by default let's always generate 10 new tokens
+        elif has_default_max_length:  # by default let's always generate 20 new tokens
             generation_config.max_length = generation_config.max_length + input_ids_length
+            max_position_embeddings = getattr(self.config, "max_position_embeddings", None)
+            if max_position_embeddings is not None:
+                generation_config.max_length = min(generation_config.max_length, max_position_embeddings)
 
         # same for min length
         if generation_config.min_new_tokens is not None:

tests/models/encoder_decoder/test_modeling_encoder_decoder.py

@@ -488,7 +488,9 @@ class EncoderDecoderMixin:
 
         # Bert does not have a bos token id, so use pad_token_id instead
         generated_output = enc_dec_model.generate(
-            input_ids, decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id
+            input_ids,
+            decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id,
+            max_length=decoder_config.max_length,
         )
         self.assertEqual(generated_output.shape, (input_ids.shape[0],) + (decoder_config.max_length,))
 

tests/models/speech_encoder_decoder/test_modeling_speech_encoder_decoder.py

@@ -362,7 +362,9 @@ class EncoderDecoderMixin:
 
         # Bert does not have a bos token id, so use pad_token_id instead
         generated_output = enc_dec_model.generate(
-            inputs, decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id
+            inputs,
+            decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id,
+            max_length=decoder_config.max_length,
         )
         self.assertEqual(generated_output.shape, (inputs.shape[0],) + (decoder_config.max_length,))
 

tests/models/vision_encoder_decoder/test_modeling_vision_encoder_decoder.py

@@ -306,7 +306,9 @@ class EncoderDecoderMixin:
 
         # Bert does not have a bos token id, so use pad_token_id instead
         generated_output = enc_dec_model.generate(
-            inputs, decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id
+            inputs,
+            decoder_start_token_id=enc_dec_model.config.decoder.pad_token_id,
+            max_length=decoder_config.max_length,
         )
         self.assertEqual(generated_output.shape, (inputs.shape[0],) + (decoder_config.max_length,))
 
@@ -873,6 +875,7 @@ class LayoutLMv32TrOCR(EncoderDecoderMixin, unittest.TestCase):
         generated_output = enc_dec_model.generate(
             pixel_values=pixel_values,
             decoder_start_token_id=enc_dec_model.config.decoder.bos_token_id,
+            max_length=decoder_config.max_length,
             **kwargs,
         )
         self.assertEqual(generated_output.shape, (pixel_values.shape[0],) + (decoder_config.max_length,))
@@ -990,6 +993,7 @@ class VIT2GPT2Test(EncoderDecoderMixin, unittest.TestCase):
         generated_output = enc_dec_model.generate(
             pixel_values=pixel_values,
             decoder_start_token_id=enc_dec_model.config.decoder.bos_token_id,
+            max_length=decoder_config.max_length,
             **kwargs,
         )
         self.assertEqual(generated_output.shape, (pixel_values.shape[0],) + (decoder_config.max_length,))
@@ -1107,6 +1111,7 @@ class Donut2GPT2Test(EncoderDecoderMixin, unittest.TestCase):
         generated_output = enc_dec_model.generate(
             pixel_values=pixel_values,
             decoder_start_token_id=enc_dec_model.config.decoder.bos_token_id,
+            max_length=decoder_config.max_length,
             **kwargs,
         )
         self.assertEqual(generated_output.shape, (pixel_values.shape[0],) + (decoder_config.max_length,))