Fix en documentation typos (#21799)

* fix wrong url

* typos in english documentation

docs/source/en/tasks/masked_language_modeling.mdx

@@ -168,7 +168,7 @@ Apply the `group_texts` function over the entire dataset:
 >>> lm_dataset = tokenized_eli5.map(group_texts, batched=True, num_proc=4)
 ```
 
-Now create a batch of examples using [`DataCollatorForLanguageModeling`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+Now create a batch of examples using [`DataCollatorForLanguageModeling`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 <frameworkcontent>
 <pt>

docs/source/en/tasks/multiple_choice.mdx

@@ -119,7 +119,7 @@ To apply the preprocessing function over the entire dataset, use 🤗 Datasets [
 tokenized_swag = swag.map(preprocess_function, batched=True)
 ```
 
-🤗 Transformers doesn't have a data collator for multiple choice, so you'll need to adapt the [`DataCollatorWithPadding`] to create a batch of examples. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+🤗 Transformers doesn't have a data collator for multiple choice, so you'll need to adapt the [`DataCollatorWithPadding`] to create a batch of examples. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 `DataCollatorForMultipleChoice` flattens all the model inputs, applies padding, and then unflattens the results:
 

docs/source/en/tasks/question_answering.mdx

@@ -54,7 +54,7 @@ We encourage you to login to your Hugging Face account so you can upload and sha
 
 ## Load SQuAD dataset
 
-Start by loading a smaller subset of the SQuAD dataset from the 🤗 Datasets library. This'll give you a chance to experiment and make sure everythings works before spending more time training on the full dataset.
+Start by loading a smaller subset of the SQuAD dataset from the 🤗 Datasets library. This'll give you a chance to experiment and make sure everything works before spending more time training on the full dataset.
 
 ```py
 >>> from datasets import load_dataset

docs/source/en/tasks/semantic_segmentation.mdx

@@ -50,7 +50,7 @@ We encourage you to log in to your Hugging Face account so you can upload and sh
 
 ## Load SceneParse150 dataset
 
-Start by loading a smaller subset of the SceneParse150 dataset from the 🤗 Datasets library. This'll give you a chance to experiment and make sure everythings works before spending more time training on the full dataset.
+Start by loading a smaller subset of the SceneParse150 dataset from the 🤗 Datasets library. This'll give you a chance to experiment and make sure everything works before spending more time training on the full dataset.
 
 ```py
 >>> from datasets import load_dataset

docs/source/en/tasks/sequence_classification.mdx

@@ -97,7 +97,7 @@ To apply the preprocessing function over the entire dataset, use 🤗 Datasets [
 tokenized_imdb = imdb.map(preprocess_function, batched=True)
 ```
 
-Now create a batch of examples using [`DataCollatorWithPadding`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+Now create a batch of examples using [`DataCollatorWithPadding`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 <frameworkcontent>
 <pt>

docs/source/en/tasks/summarization.mdx

@@ -118,7 +118,7 @@ To apply the preprocessing function over the entire dataset, use 🤗 Datasets [
 >>> tokenized_billsum = billsum.map(preprocess_function, batched=True)
 ```
 
-Now create a batch of examples using [`DataCollatorForSeq2Seq`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+Now create a batch of examples using [`DataCollatorForSeq2Seq`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 <frameworkcontent>
 <pt>

docs/source/en/tasks/token_classification.mdx

@@ -156,7 +156,7 @@ To apply the preprocessing function over the entire dataset, use 🤗 Datasets [
 >>> tokenized_wnut = wnut.map(tokenize_and_align_labels, batched=True)
 ```
 
-Now create a batch of examples using [`DataCollatorWithPadding`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+Now create a batch of examples using [`DataCollatorWithPadding`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 <frameworkcontent>
 <pt>

docs/source/en/tasks/translation.mdx

@@ -113,7 +113,7 @@ To apply the preprocessing function over the entire dataset, use 🤗 Datasets [
 >>> tokenized_books = books.map(preprocess_function, batched=True)
 ```
 
-Now create a batch of examples using [`DataCollatorForSeq2Seq`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximium length.
+Now create a batch of examples using [`DataCollatorForSeq2Seq`]. It's more efficient to *dynamically pad* the sentences to the longest length in a batch during collation, instead of padding the whole dataset to the maximum length.
 
 <frameworkcontent>
 <pt>