Remove `trust_remote_code` when loading Libri Dummy (#31748)

* [whisper integration] use parquet dataset for testing

* propagate to others

* more propagation

* last one

tests/models/clap/test_modeling_clap.py

@@ -665,9 +665,7 @@ class ClapModelIntegrationTest(unittest.TestCase):
             "repeat": 0.0023,
         }
 
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         audio_sample = librispeech_dummy[-1]
 
         model_id = "laion/clap-htsat-unfused"
@@ -694,9 +692,7 @@ class ClapModelIntegrationTest(unittest.TestCase):
             "pad": -0.000379,
         }
 
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         audio_sample = librispeech_dummy[-1]
 
         model_id = "laion/clap-htsat-fused"
@@ -723,9 +719,7 @@ class ClapModelIntegrationTest(unittest.TestCase):
             "pad": 0.0006,
         }
 
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         audio_samples = [sample["array"] for sample in librispeech_dummy[0:4]["audio"]]
 
         model_id = "laion/clap-htsat-fused"
@@ -752,9 +746,7 @@ class ClapModelIntegrationTest(unittest.TestCase):
             "pad": 0.0019,
         }
 
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         audio_samples = [sample["array"] for sample in librispeech_dummy[0:4]["audio"]]
 
         model_id = "laion/clap-htsat-unfused"

tests/models/clvp/test_feature_extraction_clvp.py

@@ -209,9 +209,7 @@ class ClvpFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unittest.Tes
             self.assertTrue(pt_processed.input_features.dtype == torch.float32)
 
     def _load_datasamples(self, num_samples):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         ds = ds.cast_column("audio", Audio(sampling_rate=22050))
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]

tests/models/clvp/test_modeling_clvp.py

@@ -371,9 +371,7 @@ class ClvpModelForConditionalGenerationTester:
     def prepare_config_and_inputs(self):
         _, input_ids, attention_mask = self.clvp_encoder_tester.prepare_config_and_inputs()
 
-        ds = datasets.load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = datasets.load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         ds = ds.cast_column("audio", datasets.Audio(sampling_rate=22050))
         _, audio, sr = ds.sort("id").select(range(1))[:1]["audio"][0].values()
 
@@ -555,9 +553,7 @@ class ClvpModelForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase)
 class ClvpIntegrationTest(unittest.TestCase):
     def setUp(self):
         self.text = "This is an example text."
-        ds = datasets.load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = datasets.load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         ds = ds.cast_column("audio", datasets.Audio(sampling_rate=22050))
         _, self.speech_samples, self.sr = ds.sort("id").select(range(1))[:1]["audio"][0].values()
 

tests/models/data2vec/test_modeling_data2vec_audio.py

@@ -694,9 +694,7 @@ class Data2VecAudioUtilsTest(unittest.TestCase):
 @slow
 class Data2VecAudioModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/encodec/test_feature_extraction_encodec.py

@@ -138,9 +138,7 @@ class EnCodecFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unittest.
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         audio_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/encodec/test_modeling_encodec.py

@@ -461,9 +461,7 @@ class EncodecIntegrationTest(unittest.TestCase):
             "1.5": [371955],
             "24.0": [6659962],
         }
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         model_id = "facebook/encodec_24khz"
 
         model = EncodecModel.from_pretrained(model_id).to(torch_device)
@@ -517,9 +515,7 @@ class EncodecIntegrationTest(unittest.TestCase):
             "3.0": [144259, 146765, 156435, 176871, 161971],
             "24.0": [1568553, 1294948, 1306190, 1464747, 1663150],
         }
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         model_id = "facebook/encodec_48khz"
 
         model = EncodecModel.from_pretrained(model_id).to(torch_device)
@@ -581,9 +577,7 @@ class EncodecIntegrationTest(unittest.TestCase):
                 [85561, 81870, 76953, 48967, 79315, 85442, 81479, 107241],
             ],
         }
-        librispeech_dummy = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        librispeech_dummy = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         model_id = "facebook/encodec_48khz"
 
         model = EncodecModel.from_pretrained(model_id).to(torch_device)

tests/models/hubert/test_modeling_hubert.py

@@ -753,9 +753,7 @@ class HubertModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/hubert/test_modeling_tf_hubert.py

@@ -609,9 +609,7 @@ class TFHubertModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/pop2piano/test_feature_extraction_pop2piano.py

@@ -136,9 +136,7 @@ class Pop2PianoFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unittes
         self.assertTrue(input_features.extrapolated_beatstep.ndim == 2)
 
     def test_integration(self):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         speech_samples = ds.sort("id").select([0])["audio"]
         input_speech = [x["array"] for x in speech_samples][0]
         sampling_rate = [x["sampling_rate"] for x in speech_samples][0]

tests/models/pop2piano/test_processor_pop2piano.py

@@ -111,9 +111,7 @@ class Pop2PianoProcessorTest(unittest.TestCase):
 
     def get_inputs(self):
         """get inputs for both feature extractor and tokenizer"""
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         speech_samples = ds.sort("id").select([0])["audio"]
         input_speech = [x["array"] for x in speech_samples][0]
         sampling_rate = [x["sampling_rate"] for x in speech_samples][0]

tests/models/seamless_m4t/test_feature_extraction_seamless_m4t.py

@@ -258,9 +258,7 @@ class SeamlessM4TFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unitt
             self.assertTrue(pt_processed.input_features.dtype == torch.float32)
 
     def _load_datasample(self, id):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_sample = ds.sort("id")[id]["audio"]["array"]
 

tests/models/sew/test_modeling_sew.py

@@ -494,9 +494,7 @@ class SEWModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/sew_d/test_modeling_sew_d.py

@@ -508,9 +508,7 @@ class SEWDModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/speech_to_text/test_feature_extraction_speech_to_text.py

@@ -259,9 +259,7 @@ class Speech2TextFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unitt
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/speech_to_text/test_modeling_speech_to_text.py

@@ -793,9 +793,7 @@ class Speech2TextModelIntegrationTests(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/speech_to_text/test_modeling_tf_speech_to_text.py

@@ -587,9 +587,7 @@ class TFSpeech2TextModelIntegrationTests(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/speecht5/test_feature_extraction_speecht5.py

@@ -380,9 +380,7 @@ class SpeechT5FeatureExtractionTest(SequenceFeatureExtractionTestMixin, unittest
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/speecht5/test_modeling_speecht5.py

@@ -744,9 +744,7 @@ class SpeechT5ForSpeechToTextIntegrationTests(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 
@@ -1771,9 +1769,7 @@ class SpeechT5ForSpeechToSpeechIntegrationTests(unittest.TestCase):
     def _load_datasamples(self, num_samples):
         from datasets import load_dataset
 
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").select(range(num_samples))[:num_samples]["audio"]
 

tests/models/unispeech/test_modeling_unispeech.py

@@ -549,9 +549,7 @@ class UniSpeechRobustModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.T
 @slow
 class UniSpeechModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]

tests/models/unispeech_sat/test_modeling_unispeech_sat.py

@@ -806,9 +806,7 @@ class UniSpeechSatRobustModelTest(ModelTesterMixin, unittest.TestCase):
 @slow
 class UniSpeechSatModelIntegrationTest(unittest.TestCase):
     def _load_datasamples(self, num_samples):
-        ds = load_dataset(
-            "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation", trust_remote_code=True
-        )
+        ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
         # automatic decoding with librispeech
         speech_samples = ds.sort("id").filter(
             lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)]