fix: change of "sentence" behaviour in DocumentSplitter (#373)

* adding fix

* adding fix

* adding fix

* fix splitter

* install nltk

* add nltk dep to index.toml

* Trigger Build

* Trigger Build

---------

Co-authored-by: anakin87 <[email protected]>

Author: davidsbatista

tutorials/39_Embedding_Metadata_for_Improved_Retrieval.ipynb

@@ -182,7 +182,7 @@
     "\n",
     "def create_indexing_pipeline(document_store, metadata_fields_to_embed=None):\n",
     "    document_cleaner = DocumentCleaner()\n",
-    "    document_splitter = DocumentSplitter(split_by=\"sentence\", split_length=2)\n",
+    "    document_splitter = DocumentSplitter(split_by=\"period\", split_length=2)\n",
     "    document_embedder = SentenceTransformersDocumentEmbedder(\n",
     "        model=\"thenlper/gte-large\", meta_fields_to_embed=metadata_fields_to_embed\n",
     "    )\n",