fix: typo spelling grammar (#13212)

* fix: typo spelling grammar

* fix: make fixup

docs/source/main_classes/trainer.rst

@@ -197,7 +197,7 @@ which should make the "stop and resume" style of training as close as possible t
 However, due to various default non-deterministic pytorch settings this might not fully work. If you want full
 determinism please refer to `Controlling sources of randomness
 <https://pytorch.org/docs/stable/notes/randomness.html>`__. As explained in the document, that some of those settings
-that make things determinstic (.e.g., ``torch.backends.cudnn.deterministic``) may slow things down, therefore this
+that make things deterministic (.e.g., ``torch.backends.cudnn.deterministic``) may slow things down, therefore this
 can't be done by default, but you can enable those yourself if needed.
 
 

docs/source/model_doc/deberta_v2.rst

@@ -53,7 +53,7 @@ New in v2:
   transformer layer to better learn the local dependency of input tokens.
 - **Sharing position projection matrix with content projection matrix in attention layer** Based on previous
   experiments, this can save parameters without affecting the performance.
-- **Apply bucket to encode relative postions** The DeBERTa-v2 model uses log bucket to encode relative positions
+- **Apply bucket to encode relative positions** The DeBERTa-v2 model uses log bucket to encode relative positions
   similar to T5.
 - **900M model & 1.5B model** Two additional model sizes are available: 900M and 1.5B, which significantly improves the
   performance of downstream tasks.

docs/source/model_doc/speech_to_text.rst

@@ -42,8 +42,8 @@ features. The :class:`~transformers.Speech2TextProcessor` wraps :class:`~transfo
 predicted token ids.
 
 The feature extractor depends on :obj:`torchaudio` and the tokenizer depends on :obj:`sentencepiece` so be sure to
-install those packages before running the examples. You could either install those as extra speech dependancies with
-``pip install transformers"[speech, sentencepiece]"`` or install the packages seperatly with ``pip install torchaudio
+install those packages before running the examples. You could either install those as extra speech dependencies with
+``pip install transformers"[speech, sentencepiece]"`` or install the packages seperately with ``pip install torchaudio
 sentencepiece``. Also ``torchaudio`` requires the development version of the `libsndfile
 <http://www.mega-nerd.com/libsndfile/>`__ package which can be installed via a system package manager. On Ubuntu it can
 be installed as follows: ``apt install libsndfile1-dev``

docs/source/training.rst

@@ -281,7 +281,7 @@ Fine-tuning in native PyTorch
    frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope;
    picture-in-picture" allowfullscreen></iframe>
 
-You might need to restart your notebook at this stage to free some memory, or excute the following code:
+You might need to restart your notebook at this stage to free some memory, or execute the following code:
 
 .. code-block:: python
 

src/transformers/deepspeed.py

@@ -62,7 +62,7 @@ class HfDeepSpeedConfig:
 
         if isinstance(config_file_or_dict, dict):
             # Don't modify user's data should they want to reuse it (e.g. in tests), because once we
-            # modified it, it will not be accepted here again, since `auto` values would have been overriden
+            # modified it, it will not be accepted here again, since `auto` values would have been overridden
             config = deepcopy(config_file_or_dict)
         elif isinstance(config_file_or_dict, str):
             with io.open(config_file_or_dict, "r", encoding="utf-8") as f:

src/transformers/modelcard.py

@@ -468,7 +468,7 @@ class TrainingSummary:
             model_card += f"This model is a fine-tuned version of [{self.finetuned_from}](https://huggingface.co/{self.finetuned_from}) on "
 
         if self.dataset is None:
-            model_card += "an unkown dataset."
+            model_card += "an unknown dataset."
         else:
             if isinstance(self.dataset, str):
                 model_card += f"the {self.dataset} dataset."

src/transformers/modeling_flax_utils.py

@@ -177,14 +177,14 @@ class FlaxPreTrainedModel(PushToHubMixin, FlaxGenerationMixin):
                     - A path or url to a `pt index checkpoint file` (e.g, ``./tf_model/model.ckpt.index``). In this
                       case, ``from_pt`` should be set to :obj:`True`.
             model_args (sequence of positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
             config (:obj:`Union[PretrainedConfig, str, os.PathLike]`, `optional`):
                 Can be either:
 
                     - an instance of a class derived from :class:`~transformers.PretrainedConfig`,
                     - a string or path valid as input to :func:`~transformers.PretrainedConfig.from_pretrained`.
 
-                Configuration for the model to use instead of an automatically loaded configuation. Configuration can
+                Configuration for the model to use instead of an automatically loaded configuration. Configuration can
                 be automatically loaded when:
 
                     - The model is a model provided by the library (loaded with the `model id` string of a pretrained

src/transformers/modeling_tf_utils.py

@@ -1120,14 +1120,14 @@ class TFPreTrainedModel(tf.keras.Model, TFModelUtilsMixin, TFGenerationMixin, Pu
                     - :obj:`None` if you are both providing the configuration and state dictionary (resp. with keyword
                       arguments ``config`` and ``state_dict``).
             model_args (sequence of positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
             config (:obj:`Union[PretrainedConfig, str]`, `optional`):
                 Can be either:
 
                     - an instance of a class derived from :class:`~transformers.PretrainedConfig`,
                     - a string valid as input to :func:`~transformers.PretrainedConfig.from_pretrained`.
 
-                Configuration for the model to use instead of an automatically loaded configuation. Configuration can
+                Configuration for the model to use instead of an automatically loaded configuration. Configuration can
                 be automatically loaded when:
 
                     - The model is a model provided by the library (loaded with the `model id` string of a pretrained

src/transformers/modeling_utils.py

@@ -1038,14 +1038,14 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
                     - :obj:`None` if you are both providing the configuration and state dictionary (resp. with keyword
                       arguments ``config`` and ``state_dict``).
             model_args (sequence of positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
             config (:obj:`Union[PretrainedConfig, str, os.PathLike]`, `optional`):
                 Can be either:
 
                     - an instance of a class derived from :class:`~transformers.PretrainedConfig`,
                     - a string or path valid as input to :func:`~transformers.PretrainedConfig.from_pretrained`.
 
-                Configuration for the model to use instead of an automatically loaded configuation. Configuration can
+                Configuration for the model to use instead of an automatically loaded configuration. Configuration can
                 be automatically loaded when:
 
                     - The model is a model provided by the library (loaded with the `model id` string of a pretrained

src/transformers/models/big_bird/modeling_big_bird.py

@@ -1138,7 +1138,7 @@ class BigBirdBlockSparseAttention(nn.Module):
             from_block_size: int. size of block in from sequence.
             to_block_size: int. size of block in to sequence.
             num_heads: int. total number of heads.
-            plan_from_length: list. plan from length where num_random_blocks are choosen from.
+            plan_from_length: list. plan from length where num_random_blocks are chosen from.
             plan_num_rand_blocks: list. number of rand blocks within the plan.
             window_block_left: int. number of blocks of window to left of a block.
             window_block_right: int. number of blocks of window to right of a block.

src/transformers/models/bigbird_pegasus/modeling_bigbird_pegasus.py

@@ -952,7 +952,7 @@ class BigBirdPegasusBlockSparseAttention(nn.Module):
             from_block_size: int. size of block in from sequence.
             to_block_size: int. size of block in to sequence.
             num_heads: int. total number of heads.
-            plan_from_length: list. plan from length where num_random_blocks are choosen from.
+            plan_from_length: list. plan from length where num_random_blocks are chosen from.
             plan_num_rand_blocks: list. number of rand blocks within the plan.
             window_block_left: int. number of blocks of window to left of a block.
             window_block_right: int. number of blocks of window to right of a block.

src/transformers/models/clip/tokenization_clip.py

@@ -60,7 +60,7 @@ def bytes_to_unicode():
 
     The reversible bpe codes work on unicode strings. This means you need a large # of unicode characters in your vocab
     if you want to avoid UNKs. When you're at something like a 10B token dataset you end up needing around 5K for
-    decent coverage. This is a signficant percentage of your normal, say, 32K bpe vocab. To avoid that, we want lookup
+    decent coverage. This is a significant percentage of your normal, say, 32K bpe vocab. To avoid that, we want lookup
     tables between utf-8 bytes and unicode strings.
     """
     bs = (
@@ -317,7 +317,7 @@ class CLIPTokenizer(PreTrainedTokenizer):
         for token in re.findall(self.pat, text):
             token = "".join(
                 self.byte_encoder[b] for b in token.encode("utf-8")
-            )  # Maps all our bytes to unicode strings, avoiding controle tokens of the BPE (spaces in our case)
+            )  # Maps all our bytes to unicode strings, avoiding control tokens of the BPE (spaces in our case)
             bpe_tokens.extend(bpe_token for bpe_token in self.bpe(token).split(" "))
         return bpe_tokens
 

src/transformers/models/detr/modeling_detr.py

@@ -151,7 +151,7 @@ class DetrObjectDetectionOutput(ModelOutput):
             unnormalized bounding boxes.
         auxiliary_outputs (:obj:`list[Dict]`, `optional`):
             Optional, only returned when auxilary losses are activated (i.e. :obj:`config.auxiliary_loss` is set to
-            `True`) and labels are provided. It is a list of dictionnaries containing the two above keys (:obj:`logits`
+            `True`) and labels are provided. It is a list of dictionaries containing the two above keys (:obj:`logits`
             and :obj:`pred_boxes`) for each decoder layer.
         last_hidden_state (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`):
             Sequence of hidden-states at the output of the last layer of the decoder of the model.
@@ -218,8 +218,8 @@ class DetrSegmentationOutput(ModelOutput):
             :meth:`~transformers.DetrFeatureExtractor.post_process_panoptic` to evaluate instance and panoptic
             segmentation masks respectively.
         auxiliary_outputs (:obj:`list[Dict]`, `optional`):
-            Optional, only returned when auxilary losses are activated (i.e. :obj:`config.auxiliary_loss` is set to
-            `True`) and labels are provided. It is a list of dictionnaries containing the two above keys (:obj:`logits`
+            Optional, only returned when auxiliary losses are activated (i.e. :obj:`config.auxiliary_loss` is set to
+            `True`) and labels are provided. It is a list of dictionaries containing the two above keys (:obj:`logits`
             and :obj:`pred_boxes`) for each decoder layer.
         last_hidden_state (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`):
             Sequence of hidden-states at the output of the last layer of the decoder of the model.

src/transformers/models/encoder_decoder/modeling_encoder_decoder.py

@@ -272,7 +272,7 @@ class EncoderDecoderModel(PreTrainedModel):
                       a PyTorch model using the provided conversion scripts and loading the PyTorch model afterwards.
 
             model_args (remaining positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
 
             kwargs (remaining dictionary of keyword arguments, `optional`):
                 Can be used to update the configuration object (after it being loaded) and initiate the model (e.g.,

src/transformers/models/gpt_neo/configuration_gpt_neo.py

@@ -205,7 +205,7 @@ def custom_unfold(input, dimension, size, step):
 def custom_get_block_length_and_num_blocks(seq_length, window_size):
     """
     Custom implementation for GPTNeoAttentionMixin._get_block_length_and_num_blocks to enable the export to ONNX as
-    original implmentation uses Python variables and control flow.
+    original implementation uses Python variables and control flow.
     """
     import torch
 

src/transformers/models/hubert/modeling_hubert.py

@@ -237,7 +237,7 @@ class HubertSamePadLayer(nn.Module):
 
 # Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2FeatureExtractor with Wav2Vec2->Hubert
 class HubertFeatureExtractor(nn.Module):
-    """Construct the featurs from raw audio waveform"""
+    """Construct the features from raw audio waveform"""
 
     def __init__(self, config):
         super().__init__()

src/transformers/models/rag/modeling_rag.py

@@ -283,7 +283,7 @@ class RagPreTrainedModel(PreTrainedModel):
                       a PyTorch model using the provided conversion scripts and loading the PyTorch model afterwards.
 
             model_args (remaining positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
             retriever (:class:`~transformers.RagRetriever`, `optional`):
                 The retriever to use.
             kwwargs (remaining dictionary of keyword arguments, `optional`):

src/transformers/models/rag/modeling_tf_rag.py

@@ -258,7 +258,7 @@ class TFRagPreTrainedModel(TFPreTrainedModel):
                       ``generator_from_pt`` should be set to :obj:`True`.
 
             model_args (remaining positional arguments, `optional`):
-                All remaning positional arguments will be passed to the underlying model's ``__init__`` method.
+                All remaining positional arguments will be passed to the underlying model's ``__init__`` method.
             retriever (:class:`~transformers.RagRetriever`, `optional`):
                 The retriever to use.
             kwargs (remaining dictionary of keyword arguments, `optional`):

src/transformers/models/wav2vec2/modeling_flax_wav2vec2.py

@@ -385,7 +385,7 @@ class FlaxConvLayersCollection(nn.Module):
 
 
 class FlaxWav2Vec2FeatureExtractor(nn.Module):
-    """Construct the featurs from raw audio waveform"""
+    """Construct the features from raw audio waveform"""
 
     config: Wav2Vec2Config
     dtype: jnp.dtype = jnp.float32

src/transformers/models/wav2vec2/modeling_wav2vec2.py

@@ -308,7 +308,7 @@ class Wav2Vec2SamePadLayer(nn.Module):
 
 
 class Wav2Vec2FeatureExtractor(nn.Module):
-    """Construct the featurs from raw audio waveform"""
+    """Construct the features from raw audio waveform"""
 
     def __init__(self, config):
         super().__init__()

src/transformers/onnx/convert.py

@@ -158,7 +158,7 @@ def validate_model_outputs(
 
     # We flatten potential collection of outputs (i.e. past_keys) to a flat structure
     for name, value in ref_outputs.items():
-        # Overwriting the output name as "present" since it is the name used for the ONNX ouputs
+        # Overwriting the output name as "present" since it is the name used for the ONNX outputs
         # ("past_key_values" being taken for the ONNX inputs)
         if name == "past_key_values":
             name = "present"

src/transformers/onnx/features.py

@@ -114,7 +114,7 @@ class FeaturesManager:
 
         Args:
             model: The model to export
-            feature: The name of the feature to check if it is avaiable
+            feature: The name of the feature to check if it is available
 
         Returns:
             (str) The type of the model (OnnxConfig) The OnnxConfig instance holding the model export properties

src/transformers/tokenization_utils_base.py

@@ -1375,7 +1375,7 @@ INIT_TOKENIZER_DOCSTRING = r"""
           high-level keys being the ``__init__`` keyword name of each vocabulary file required by the model, the
           low-level being the :obj:`short-cut-names` of the pretrained models with, as associated values, the
           :obj:`url` to the associated pretrained vocabulary file.
-        - **max_model_input_sizes** (:obj:`Dict[str, Optinal[int]]`) -- A dictionary with, as keys, the
+        - **max_model_input_sizes** (:obj:`Dict[str, Optional[int]]`) -- A dictionary with, as keys, the
           :obj:`short-cut-names` of the pretrained models, and as associated values, the maximum length of the sequence
           inputs of this model, or :obj:`None` if the model has no maximum input size.
         - **pretrained_init_configuration** (:obj:`Dict[str, Dict[str, Any]]`) -- A dictionary with, as keys, the
@@ -1785,7 +1785,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
                 config = AutoConfig.from_pretrained(pretrained_model_name_or_path)
                 config_tokenizer_class = config.tokenizer_class
             except (OSError, ValueError, KeyError):
-                # skip if an error occured.
+                # skip if an error occurred.
                 config = None
             if config_tokenizer_class is None:
                 # Third attempt. If we have not yet found the original type of the tokenizer,

src/transformers/tokenization_utils_fast.py

@@ -707,7 +707,7 @@ class PreTrainedTokenizerFast(PreTrainedTokenizerBase):
 
                 special_token_full = getattr(self, f"_{token}")
                 if isinstance(special_token_full, AddedToken):
-                    # Create an added token with the same paramters except the content
+                    # Create an added token with the same parameters except the content
                     kwargs[token] = AddedToken(
                         special_token,
                         single_word=special_token_full.single_word,