Update README.md

Fixing more typos.

README.md

@@ -22,12 +22,12 @@ The main data is organized in a JSON file — [questions_QA_2024_v1.0.json](ques
 * `question_summary`: A very brief description of the question subject.
 * `question_text`: The question itself.
 * `answer`: Answer, as extracted from the original document. It is a list of strings, respecting the PDF formatting, containing all the information provided after the `question_text` and before a link to the document index, provided at the end of all questions.
-* `answer_cleaned`: The `answer` field **after removing** all *explicit external references* (references to official legal documents) and all *explicit internal references* (references to other questions) ― external references were provided in the original document in two ways: explicitly, through *greyed* boxes, implicitly, embedded in the answer text.
+* `answer_cleaned`: The `answer` field **after removing** all *explicit external references* (references to official legal documents) and all *explicit internal references* (references to other questions) ― external references were provided in the original document in two ways: explicitly, through *grayed* boxes, implicitly, embedded in the answer text.
 * `references`: External references **explicitly** provided.
 * `linked_questions`: Other questions referred in the provided answer.
 * `formatted_references`: The explicit external references, LLM-processed to separate the document title, articles, sections, paragraphs and other specific parts mentioned.
 * `embedded_references`: External references **implicitly** provided, embedded in the answer text.
-* `formatted_embedded_references`: The implicit external references LLM-processed to separated the specific information mentioned, similar to the `formatted_references` field.
+* `formatted_embedded_references`: The implicit external references LLM-processed to separate the specific information mentioned, similar to the `formatted_references` field.
 * `all_formated_references`: Merge of `formatted_references` and `formatted_embedded_references` fields, combining the information of the legal documents, and including the name of the text file (the `file` sub-field) were each particular legal document has been captured in the dataset.