Fix some typos. (#17560)

* Fix some typos.
Signed-off-by: Yulv-git <yulvchi@qq.com>

* Fix typo.
Signed-off-by: Yulv-git <yulvchi@qq.com>

* make fixup.

src/transformers/models/hubert/modeling_hubert.py

@@ -174,7 +174,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length

src/transformers/models/hubert/modeling_tf_hubert.py

@@ -203,7 +203,7 @@ def _compute_mask_indices(
     Computes random mask spans for a given shape
 
     Args:
-        shape: the the shape for which to compute masks.
+        shape: the shape for which to compute masks.
             should be of size 2 where first element is batch size and 2nd is timesteps
         attention_mask: optional padding mask of the same size as shape, which will prevent masking padded elements
         mask_prob:

src/transformers/models/led/modeling_tf_led.py

@@ -2330,7 +2330,7 @@ class TFLEDForConditionalGeneration(TFLEDPreTrainedModel):
         super().__init__(config, *inputs, **kwargs)
         self.led = TFLEDMainLayer(config, name="led")
         self.use_cache = config.use_cache
-        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the the sake of consistency.
+        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the sake of consistency.
         self.final_logits_bias = self.add_weight(
             name="final_logits_bias", shape=[1, config.vocab_size], initializer="zeros", trainable=False
         )

src/transformers/models/lxmert/modeling_lxmert.py

@@ -1110,7 +1110,7 @@ class LxmertForPreTraining(LxmertPreTrainedModel):
 
     def get_qa_logit_layer(self) -> nn.Module:
         """
-        Returns the the linear layer that produces question answering logits.
+        Returns the linear layer that produces question answering logits.
 
         Returns:
             `nn.Module`: A torch module mapping the question answering prediction hidden states or `None` if LXMERT
@@ -1341,7 +1341,7 @@ class LxmertForQuestionAnswering(LxmertPreTrainedModel):
 
     def get_qa_logit_layer(self) -> nn.Module:
         """
-        Returns the the linear layer that produces question answering logits
+        Returns the linear layer that produces question answering logits
 
         Returns:
             `nn.Module`: A torch module mapping the question answering prediction hidden states. `None`: A NoneType

src/transformers/models/marian/modeling_tf_marian.py

@@ -1283,7 +1283,7 @@ class TFMarianMTModel(TFMarianPreTrainedModel, TFCausalLanguageModelingLoss):
         super().__init__(config, *inputs, **kwargs)
         self.model = TFMarianMainLayer(config, name="model")
         self.use_cache = config.use_cache
-        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the the sake of consistency.
+        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the sake of consistency.
         self.final_logits_bias = self.add_weight(
             name="final_logits_bias", shape=[1, config.vocab_size], initializer="zeros", trainable=False
         )

src/transformers/models/maskformer/modeling_maskformer.py

@@ -1912,7 +1912,7 @@ class MaskFormerLoss(nn.Module):
 
     def get_num_masks(self, class_labels: torch.Tensor, device: torch.device) -> torch.Tensor:
         """
-        Computes the average number of target masks accross the batch, for normalization purposes.
+        Computes the average number of target masks across the batch, for normalization purposes.
         """
         num_masks = sum([len(classes) for classes in class_labels])
         num_masks_pt = torch.as_tensor([num_masks], dtype=torch.float, device=device)

src/transformers/models/mbart/modeling_tf_mbart.py

@@ -1280,7 +1280,7 @@ class TFMBartForConditionalGeneration(TFMBartPreTrainedModel, TFCausalLanguageMo
         super().__init__(config, *inputs, **kwargs)
         self.model = TFMBartMainLayer(config, name="model")
         self.use_cache = config.use_cache
-        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the the sake of consistency.
+        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the sake of consistency.
         self.final_logits_bias = self.add_weight(
             name="final_logits_bias", shape=[1, config.vocab_size], initializer="zeros", trainable=False
         )

src/transformers/models/pegasus/modeling_tf_pegasus.py

@@ -1292,7 +1292,7 @@ class TFPegasusForConditionalGeneration(TFPegasusPreTrainedModel, TFCausalLangua
         super().__init__(config, *inputs, **kwargs)
         self.model = TFPegasusMainLayer(config, name="model")
         self.use_cache = config.use_cache
-        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the the sake of consistency.
+        # final_bias_logits is registered as a buffer in pytorch, so not trainable for the sake of consistency.
         self.final_logits_bias = self.add_weight(
             name="final_logits_bias", shape=[1, config.vocab_size], initializer="zeros", trainable=False
         )

src/transformers/models/rag/configuration_rag.py

@@ -28,7 +28,7 @@ RAG_CONFIG_DOC = r"""
         title_sep (`str`, *optional*, defaults to  `" / "`):
             Separator inserted between the title and the text of the retrieved document when calling [`RagRetriever`].
         doc_sep (`str`, *optional*, defaults to  `" // "`):
-            Separator inserted between the the text of the retrieved document and the original input when calling
+            Separator inserted between the text of the retrieved document and the original input when calling
             [`RagRetriever`].
         n_docs (`int`, *optional*, defaults to 5):
             Number of documents to retrieve.

src/transformers/models/sew/configuration_sew.py

@@ -81,10 +81,10 @@ class SEWConfig(PretrainedConfig):
             feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers.
         conv_stride (`Tuple[int]` or `List[int]`, *optional*, defaults to `(5, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1)`):
             A tuple of integers defining the stride of each 1D convolutional layer in the feature encoder. The length
-            of *conv_stride* defines the number of convolutional layers and has to match the the length of *conv_dim*.
+            of *conv_stride* defines the number of convolutional layers and has to match the length of *conv_dim*.
         conv_kernel (`Tuple[int]` or `List[int]`, *optional*, defaults to `(10, 3, 1, 3, 1, 3, 1, 3, 1, 2, 1, 2, 1)`):
             A tuple of integers defining the kernel size of each 1D convolutional layer in the feature encoder. The
-            length of *conv_kernel* defines the number of convolutional layers and has to match the the length of
+            length of *conv_kernel* defines the number of convolutional layers and has to match the length of
             *conv_dim*.
         conv_bias (`bool`, *optional*, defaults to `False`):
             Whether the 1D convolutional layers have a bias.

src/transformers/models/sew/modeling_sew.py

@@ -174,7 +174,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length

src/transformers/models/sew_d/configuration_sew_d.py

@@ -99,10 +99,10 @@ class SEWDConfig(PretrainedConfig):
             feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers.
         conv_stride (`Tuple[int]` or `List[int]`, *optional*, defaults to `(5, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1)`):
             A tuple of integers defining the stride of each 1D convolutional layer in the feature encoder. The length
-            of *conv_stride* defines the number of convolutional layers and has to match the the length of *conv_dim*.
+            of *conv_stride* defines the number of convolutional layers and has to match the length of *conv_dim*.
         conv_kernel (`Tuple[int]` or `List[int]`, *optional*, defaults to `(10, 3, 1, 3, 1, 3, 1, 3, 1, 2, 1, 2, 1)`):
             A tuple of integers defining the kernel size of each 1D convolutional layer in the feature encoder. The
-            length of *conv_kernel* defines the number of convolutional layers and has to match the the length of
+            length of *conv_kernel* defines the number of convolutional layers and has to match the length of
             *conv_dim*.
         conv_bias (`bool`, *optional*, defaults to `False`):
             Whether the 1D convolutional layers have a bias.

src/transformers/models/sew_d/modeling_sew_d.py

@@ -175,7 +175,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length

src/transformers/models/unispeech/configuration_unispeech.py

@@ -85,10 +85,10 @@ class UniSpeechConfig(PretrainedConfig):
             feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers.
         conv_stride (`Tuple[int]` or `List[int]`, *optional*, defaults to `(5, 2, 2, 2, 2, 2, 2)`):
             A tuple of integers defining the stride of each 1D convolutional layer in the feature encoder. The length
-            of *conv_stride* defines the number of convolutional layers and has to match the the length of *conv_dim*.
+            of *conv_stride* defines the number of convolutional layers and has to match the length of *conv_dim*.
         conv_kernel (`Tuple[int]` or `List[int]`, *optional*, defaults to `(10, 3, 3, 3, 3, 3, 3)`):
             A tuple of integers defining the kernel size of each 1D convolutional layer in the feature encoder. The
-            length of *conv_kernel* defines the number of convolutional layers and has to match the the length of
+            length of *conv_kernel* defines the number of convolutional layers and has to match the length of
             *conv_dim*.
         conv_bias (`bool`, *optional*, defaults to `False`):
             Whether the 1D convolutional layers have a bias.

src/transformers/models/unispeech/modeling_unispeech.py

@@ -210,7 +210,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length

src/transformers/models/unispeech_sat/configuration_unispeech_sat.py

@@ -86,10 +86,10 @@ class UniSpeechSatConfig(PretrainedConfig):
             feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers.
         conv_stride (`Tuple[int]` or `List[int]`, *optional*, defaults to `(5, 2, 2, 2, 2, 2, 2)`):
             A tuple of integers defining the stride of each 1D convolutional layer in the feature encoder. The length
-            of *conv_stride* defines the number of convolutional layers and has to match the the length of *conv_dim*.
+            of *conv_stride* defines the number of convolutional layers and has to match the length of *conv_dim*.
         conv_kernel (`Tuple[int]` or `List[int]`, *optional*, defaults to `(10, 3, 3, 3, 3, 3, 3)`):
             A tuple of integers defining the kernel size of each 1D convolutional layer in the feature encoder. The
-            length of *conv_kernel* defines the number of convolutional layers and has to match the the length of
+            length of *conv_kernel* defines the number of convolutional layers and has to match the length of
             *conv_dim*.
         conv_bias (`bool`, *optional*, defaults to `False`):
             Whether the 1D convolutional layers have a bias.

src/transformers/models/unispeech_sat/modeling_unispeech_sat.py

@@ -224,7 +224,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length

src/transformers/models/wav2vec2/modeling_flax_wav2vec2.py

@@ -120,7 +120,7 @@ def _compute_mask_indices(
     CPU as part of the preprocessing during training.
 
     Args:
-        shape: the the shape for which to compute masks.
+        shape: the shape for which to compute masks.
             should be of size 2 where first element is batch size and 2nd is timesteps
         mask_prob:
             probability for each token to be chosen as start of the span to be masked. this will be multiplied by

src/transformers/models/wav2vec2/modeling_tf_wav2vec2.py

@@ -244,7 +244,7 @@ def _compute_mask_indices(
     Computes random mask spans for a given shape
 
     Args:
-        shape: the the shape for which to compute masks.
+        shape: the shape for which to compute masks.
             should be of size 2 where first element is batch size and 2nd is timesteps
         attention_mask: optional padding mask of the same size as shape, which will prevent masking padded elements
         mask_prob:

src/transformers/models/wav2vec2/modeling_wav2vec2.py

@@ -234,7 +234,7 @@ def _compute_mask_indices(
     )
     spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length)
 
-    # add offset to the starting indexes so that that indexes now create a span
+    # add offset to the starting indexes so that indexes now create a span
     offsets = np.arange(mask_length)[None, None, :]
     offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape(
         batch_size, max_num_masked_span * mask_length