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.

docs/source/en/custom_models.mdx

@@ -289,7 +289,7 @@ from huggingface_hub import notebook_login
 notebook_login()
 ```
 
-You can then push to to your own namespace (or an organization you are a member of) like this:
+You can then push to your own namespace (or an organization you are a member of) like this:
 
 ```py
 resnet50d.push_to_hub("custom-resnet50d")

docs/source/en/model_doc/speech_to_text.mdx

@@ -37,7 +37,7 @@ predicted token ids.
 
 The feature extractor depends on `torchaudio` and the tokenizer depends on `sentencepiece` so be sure to
 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
+`pip install transformers"[speech, sentencepiece]"` or install the packages separately 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/en/testing.mdx

@@ -1226,7 +1226,7 @@ This whole process would have been much easier if we only could set something li
 experimental step, and let it fail without impacting the overall status of PRs. But as mentioned earlier CircleCI and
 Github Actions don't support it at the moment.
 
-You can vote for this feature and see where it is at at these CI-specific threads:
+You can vote for this feature and see where it is at these CI-specific threads:
 
 - [Github Actions:](https://github.com/actions/toolkit/issues/399)
 - [CircleCI:](https://ideas.circleci.com/ideas/CCI-I-344)

examples/legacy/token-classification/utils_ner.py

@@ -140,7 +140,7 @@ class TokenClassificationTask:
             # it easier for the model to learn the concept of sequences.
             #
             # For classification tasks, the first vector (corresponding to [CLS]) is
-            # used as as the "sentence vector". Note that this only makes sense because
+            # used as the "sentence vector". Note that this only makes sense because
             # the entire model is fine-tuned.
             tokens += [sep_token]
             label_ids += [pad_token_label_id]

examples/research_projects/codeparrot/README.md

@@ -91,7 +91,7 @@ python scripts/initialize_model.py \
 --model_name codeparrot \
 --push_to_hub True
 ```
-This will initialize a new model with the architecture and configuration of `gpt2-large` and use the tokenizer to appropriately size the input embeddings. Finally, the initilaized model is pushed the the hub.
+This will initialize a new model with the architecture and configuration of `gpt2-large` and use the tokenizer to appropriately size the input embeddings. Finally, the initilaized model is pushed the hub.
 
 We can either pass the name of a text dataset or a pretokenized dataset which speeds up training a bit.
 Now that the tokenizer and model are also ready we can start training the model. The main training script is built with `accelerate` to scale across a wide range of platforms and infrastructure scales. We train two models with [110M](https://huggingface.co/lvwerra/codeparrot-small/) and [1.5B](https://huggingface.co/lvwerra/codeparrot/) parameters for 25-30B tokens on a 16xA100 (40GB) machine which takes 1 day and 1 week, respectively.

examples/research_projects/distillation/scripts/token_counts.py

@@ -43,7 +43,7 @@ if __name__ == "__main__":
     with open(args.data_file, "rb") as fp:
         data = pickle.load(fp)
 
-    logger.info("Counting occurences for MLM.")
+    logger.info("Counting occurrences for MLM.")
     counter = Counter()
     for tk_ids in data:
         counter.update(tk_ids)

examples/research_projects/jax-projects/big_bird/evaluate.py

@@ -106,7 +106,7 @@ def main():
         return start_logits, end_logits, jnp.argmax(pooled_logits, axis=-1)
 
     def evaluate(example):
-        # encode question and context so that they are seperated by a tokenizer.sep_token and cut at max_length
+        # encode question and context so that they are separated by a tokenizer.sep_token and cut at max_length
         inputs = tokenizer(
             example["question"],
             example["context"],

examples/research_projects/jax-projects/model_parallel/README.md

@@ -22,7 +22,7 @@ the JAX/Flax backend and the [`pjit`](https://jax.readthedocs.io/en/latest/jax.e
 > Note: The example is experimental and might have bugs. Also currently it only supports single V3-8.
 
 The `partition.py` file defines the `PyTree` of `ParitionSpec` for the GPTNeo model which describes how the model will be sharded.
-The actual sharding is auto-matically handled by `pjit`. The weights are sharded accross all local devices.
+The actual sharding is auto-matically handled by `pjit`. The weights are sharded across all local devices.
 To adapt the script for other models, we need to also change the `ParitionSpec` accordingly.
 
 TODO: Add more explantion.

src/transformers/generation_flax_utils.py

@@ -78,7 +78,7 @@ class FlaxBeamSearchOutput(ModelOutput):
         sequences (`jnp.ndarray` of shape `(batch_size, max_length)`):
             The generated sequences.
         scores (`jnp.ndarray` of shape `(batch_size,)`):
-            The scores (log probabilites) of the generated sequences.
+            The scores (log probabilities) of the generated sequences.
     """
 
     sequences: jnp.ndarray = None

src/transformers/keras_callbacks.py

@@ -277,7 +277,7 @@ class PushToHubCallback(Callback):
             for instance `"user_name/model"`, which allows you to push to an organization you are a member of with
             `"organization_name/model"`.
 
-            Will default to to the name of `output_dir`.
+            Will default to the name of `output_dir`.
         hub_token (`str`, *optional*):
             The token to use to push the model to the Hub. Will default to the token in the cache folder obtained with
             `huggingface-cli login`.

src/transformers/models/bart/modeling_tf_bart.py

@@ -1267,7 +1267,7 @@ class TFBartForConditionalGeneration(TFBartPretrainedModel, TFCausalLanguageMode
         super().__init__(config, *inputs, **kwargs)
         self.model = TFBartMainLayer(config, load_weight_prefix=load_weight_prefix, 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/blenderbot/modeling_tf_blenderbot.py

@@ -1253,7 +1253,7 @@ class TFBlenderbotForConditionalGeneration(TFBlenderbotPreTrainedModel, TFCausal
         super().__init__(config, *inputs, **kwargs)
         self.model = TFBlenderbotMainLayer(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/blenderbot_small/modeling_tf_blenderbot_small.py

@@ -1240,7 +1240,7 @@ class TFBlenderbotSmallForConditionalGeneration(TFBlenderbotSmallPreTrainedModel
         super().__init__(config, *inputs, **kwargs)
         self.model = TFBlenderbotSmallMainLayer(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/data2vec/modeling_data2vec_audio.py

@@ -184,7 +184,7 @@ 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/detr/modeling_detr.py

@@ -2054,7 +2054,7 @@ class DetrLoss(nn.Module):
         # Retrieve the matching between the outputs of the last layer and the targets
         indices = self.matcher(outputs_without_aux, targets)
 
-        # Compute the average number of target boxes accross all nodes, for normalization purposes
+        # Compute the average number of target boxes across all nodes, for normalization purposes
         num_boxes = sum(len(t["class_labels"]) for t in targets)
         num_boxes = torch.as_tensor([num_boxes], dtype=torch.float, device=next(iter(outputs.values())).device)
         # (Niels): comment out function below, distributed training to be added

src/transformers/models/electra/modeling_tf_electra.py

@@ -212,7 +212,7 @@ class TFElectraSelfOutput(tf.keras.layers.Layer):
         return hidden_states
 
 
-# Copied from from transformers.models.bert.modeling_tf_bert.TFBertAttention with Bert->Electra
+# Copied from transformers.models.bert.modeling_tf_bert.TFBertAttention with Bert->Electra
 class TFElectraAttention(tf.keras.layers.Layer):
     def __init__(self, config: ElectraConfig, **kwargs):
         super().__init__(**kwargs)

src/transformers/models/hubert/configuration_hubert.py

@@ -82,10 +82,10 @@ class HubertConfig(PretrainedConfig):
             feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers.
         conv_stride (`Tuple[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]`, *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.