[generate] fix breaking change for patch (#29976)

* fix bug and add tests

* nit

* otherway to get the cur len instead of attention mask

* more places where this might have been broken

* nit

* oups

* inputs_embeds vs input_embeds

* test generated outptus

* style

* nit

* fix

* skip failing biogpt

tests/models/biogpt/test_modeling_biogpt.py

@@ -414,6 +414,10 @@ class BioGptModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMix
         result = model(input_ids, attention_mask=attention_mask, labels=sequence_labels)
         self.assertEqual(result.logits.shape, (self.model_tester.batch_size, self.model_tester.num_labels))
 
+    @unittest.skip("The `input_embeds` when fed don't produce the same results.")
+    def test_beam_sample_generate(self):
+        pass
+
 
 @require_torch
 class BioGptModelIntegrationTest(unittest.TestCase):