For English language there are libraries like NLTK, CoreNLP which are used for Text Normalization, Word Tokenization and Detokenization, Sentence Splitting etc. Like English, is there any library to do above operation using Hindi Script ?
There is an Indic NLP library developed by iitB for Hindi texts. you can check out the below links
There're few pretrained embedders i.e. LASER, which covers Hindi as well.
If you want to achieve seminal-like scores, I wouldn't bother doing this all by hand, and take on full data-driven approach.