[WHISPER] Small patch (#21307)

* add small patch

* update tests, forced decoder ids is not prioritary against generation config

* fix two new tests

src/transformers/generation/logits_process.py

@@ -936,18 +936,19 @@ class WhisperTimeStampLogitsProcessor(LogitsProcessor):
         self.no_timestamps_token_id = generate_config.no_timestamps_token_id
         self.timestamp_begin = generate_config.no_timestamps_token_id + 1
 
-        self.begin_index = len(generate_config.forced_decoder_ids) + 1
+        self.begin_index = len(generate_config.forced_decoder_ids) + 2
         if generate_config.forced_decoder_ids[-1][1] == self.no_timestamps_token_id:
             self.begin_index -= 1
-        if generate_config.is_multilingual:
-            self.begin_index += 1
         self.max_initial_timestamp_index = generate_config.max_initial_timestamp_index
 
     def __call__(self, input_ids, scores):
         # suppress <|notimestamps|> which is handled by without_timestamps
         scores[:, self.no_timestamps_token_id] = -float("inf")
-        if input_ids.shape[1] == self.begin_index:
+
+        if input_ids.shape[1] == self.begin_index - 1:
+            scores[:, :] = -float("inf")
             scores[:, self.timestamp_begin] = 0
+            return scores
 
         # timestamps have to appear in pairs, except directly before eos_token; mask logits accordingly
         for k in range(input_ids.shape[0]):

tests/models/whisper/test_modeling_tf_whisper.py

@@ -699,8 +699,9 @@ def _test_large_generation(in_queue, out_queue, timeout):
         input_speech = _load_datasamples(1)
         input_features = processor.feature_extractor(raw_speech=input_speech, return_tensors="tf").input_features
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="en", task="transcribe")
-        generated_ids = model.generate(input_features, do_sample=False, max_length=20)
+        generated_ids = model.generate(
+            input_features, do_sample=False, max_length=20, language="<|en|>", task="transcribe"
+        )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = " Mr. Quilter is the apostle of the middle classes and we are glad"
@@ -728,26 +729,25 @@ def _test_large_generation_multilingual(in_queue, out_queue, timeout):
         input_speech = next(iter(ds))["audio"]["array"]
         input_features = processor.feature_extractor(raw_speech=input_speech, return_tensors="tf").input_features
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="ja", task="transcribe")
-        generated_ids = model.generate(input_features, do_sample=False, max_length=20)
+        generated_ids = model.generate(
+            input_features, do_sample=False, max_length=20, language="<|ja|>", task="transcribe"
+        )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = "木村さんに電話を貸してもらいました"
         unittest.TestCase().assertEqual(transcript, EXPECTED_TRANSCRIPT)
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="en", task="transcribe")
         generated_ids = model.generate(
-            input_features,
-            do_sample=False,
-            max_length=20,
+            input_features, do_sample=False, max_length=20, language="<|en|>", task="transcribe"
         )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = " Kimura-san called me."
         unittest.TestCase().assertEqual(transcript, EXPECTED_TRANSCRIPT)
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="ja", task="translate")
-        generated_ids = model.generate(input_features, do_sample=False, max_length=20)
+        generated_ids = model.generate(
+            input_features, do_sample=False, max_length=20, language="<|ja|>", task="translate"
+        )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = " I borrowed a phone from Kimura san"
@@ -779,10 +779,10 @@ def _test_large_batched_generation(in_queue, out_queue, timeout):
         # fmt: off
         EXPECTED_LOGITS = tf.convert_to_tensor(
             [
-                [50258, 50358, 50363, 2221, 13, 2326, 388, 391, 307, 264, 50244, 295, 264, 2808, 5359, 293, 321, 366, 5404, 281],
-                [50258, 50358, 50363, 6966, 307, 2221, 13, 2326, 388, 391, 311, 9060, 1570, 1880, 813, 702, 1871, 13, 50257, 50257],
-                [50258, 50358, 50363, 634, 5112, 505, 300, 412, 341, 42729, 3196, 295, 264, 1064, 11, 365, 5272, 293, 12904, 9256],
-                [50258, 50358, 50363, 634, 575, 12525, 22618, 1968, 6144, 35617, 20084, 1756, 311, 589, 307, 534, 10281, 934, 439, 11]
+                [50258, 50259, 50358, 50363, 2221, 13, 2326, 388, 391, 307, 264, 50244, 295, 264, 2808, 5359, 293, 321, 366, 5404],
+                [50258, 50259, 50358, 50363, 6966, 307, 2221, 13, 2326, 388, 391, 311, 9060, 1570, 1880, 813, 702, 1871, 13, 50257],
+                [50258, 50259, 50358, 50363, 634, 5112, 505, 300, 412, 341, 42729, 3196, 295, 264, 1064, 11, 365, 5272, 293, 12904],
+                [50258, 50259, 50358, 50363, 634, 575, 12525, 22618, 1968, 6144, 35617, 20084, 1756, 311, 589, 307, 534, 10281, 934, 439]
             ]
         )
         # fmt: on
@@ -791,10 +791,10 @@ def _test_large_batched_generation(in_queue, out_queue, timeout):
 
         # fmt: off
         EXPECTED_TRANSCRIPT = [
-            ' Mr. Quilter is the apostle of the middle classes and we are glad to',
+            " Mr. Quilter is the apostle of the middle classes and we are glad",
             " Nor is Mr. Quilter's manner less interesting than his matter.",
-            " He tells us that at this festive season of the year, with Christmas and roast beef",
-            " He has grave doubts whether Sir Frederick Layton's work is really Greek after all,"
+            " He tells us that at this festive season of the year, with Christmas and roast",
+            " He has grave doubts whether Sir Frederick Layton's work is really Greek after all",
         ]
         # fmt: on
 

tests/models/whisper/test_modeling_whisper.py

@@ -945,11 +945,8 @@ class WhisperModelIntegrationTests(unittest.TestCase):
             torch_device
         )
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="en", task="transcribe")
         generated_ids = model.generate(
-            input_features,
-            do_sample=False,
-            max_length=20,
+            input_features, do_sample=False, max_length=20, language="<|en|>", task="transcribe"
         )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
@@ -971,26 +968,25 @@ class WhisperModelIntegrationTests(unittest.TestCase):
             torch_device
         )
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="ja", task="transcribe")
-        generated_ids = model.generate(input_features, do_sample=False, max_length=20)
+        generated_ids = model.generate(
+            input_features, do_sample=False, max_length=20, language="<|ja|>", task="transcribe"
+        )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = "木村さんに電話を貸してもらいました"
         self.assertEqual(transcript, EXPECTED_TRANSCRIPT)
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="en", task="transcribe")
         generated_ids = model.generate(
-            input_features,
-            do_sample=False,
-            max_length=20,
+            input_features, do_sample=False, max_length=20, language="<|en|>", task="transcribe"
         )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = " Kimura-san called me."
         self.assertEqual(transcript, EXPECTED_TRANSCRIPT)
 
-        model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="ja", task="translate")
-        generated_ids = model.generate(input_features, do_sample=False, max_length=20)
+        generated_ids = model.generate(
+            input_features, do_sample=False, max_length=20, language="<|ja|>", task="translate"
+        )
         transcript = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
 
         EXPECTED_TRANSCRIPT = " I borrowed a phone from Kimura san"
@@ -1009,10 +1005,10 @@ class WhisperModelIntegrationTests(unittest.TestCase):
         # fmt: off
         EXPECTED_LOGITS = torch.tensor(
             [
-                [50258, 50358, 50363, 2221, 13, 2326, 388, 391, 307, 264, 50244, 295, 264, 2808, 5359, 293, 321, 366, 5404, 281],
-                [50258, 50358, 50363, 6966, 307, 2221, 13, 2326, 388, 391, 311, 9060, 1570, 1880, 813, 702, 1871, 13, 50257, 50257],
-                [50258, 50358, 50363, 634, 5112, 505, 300, 412, 341, 42729, 3196, 295, 264, 1064, 11, 365, 5272, 293, 12904, 9256],
-                [50258, 50358, 50363, 634, 575, 12525, 22618, 1968, 6144, 35617, 20084, 1756, 311, 589, 307, 534, 10281, 934, 439, 11]
+                [50258, 50259, 50358, 50363, 2221, 13, 2326, 388, 391, 307, 264, 50244, 295, 264, 2808, 5359, 293, 321, 366, 5404],
+                [50258, 50259, 50358, 50363, 6966, 307, 2221, 13, 2326, 388, 391, 311, 9060, 1570, 1880, 813, 702, 1871, 13, 50257],
+                [50258, 50259, 50358, 50363, 634, 5112, 505, 300, 412, 341, 42729, 3196, 295, 264, 1064, 11, 365, 5272, 293, 12904],
+                [50258, 50259, 50358, 50363, 634, 575, 12525, 22618, 1968, 6144, 35617, 20084, 1756, 311, 589, 307, 534, 10281, 934, 439]
             ]
         )
         # fmt: on
@@ -1021,10 +1017,10 @@ class WhisperModelIntegrationTests(unittest.TestCase):
 
         # fmt: off
         EXPECTED_TRANSCRIPT = [
-            " Mr. Quilter is the apostle of the middle classes and we are glad to",
+            " Mr. Quilter is the apostle of the middle classes and we are glad",
             " Nor is Mr. Quilter's manner less interesting than his matter.",
-            " He tells us that at this festive season of the year, with Christmas and roast beef",
-            " He has grave doubts whether Sir Frederick Layton's work is really Greek after all,",
+            " He tells us that at this festive season of the year, with Christmas and roast",
+            " He has grave doubts whether Sir Frederick Layton's work is really Greek after all",
         ]
         # fmt: on