Ungreedy subword tokenizer and vocabulary trainer for Python, Go & Javascript

A Python tool for splitting large Markdown files into smaller sections based on a specified token limit. This is particularly useful for processing large Markdown files with GPT models, as it allows the models to handle the data in manageable chunks.

Successfully developed a resume classification model which can accurately classify the resume of any person into its corresponding job with a tremendously high accuracy of more than 99%.

Machine Learning & Natural Language Processing: Reads Classic Novels and Predicts the Author of a Phrase

Deep learning-based subtitle generation model that processes audio datasets to generate accurate text transcriptions. Includes audio feature extraction, encoder-decoder architecture, training pipelines, and evaluation metrics for subtitle alignment.

Kingchop ⚔️ is a JavaScript English based library for tokenizing text (chopping text). It uses vast rules for tokenizing, and you can adjust them easily.

The tok function is a JavaScript and Node.js function that processes object instances and tokenizes text arrays. It returns tokenized words number, tokenized words array, and tokenized words concatenated string. It's part of the open-source DropSuit NLP library under the Apache License 2.0.

Nihotip is a web app that lets users explore Japanese text through interactive tokenization and detailed insights. Built with React and Python, it offers a dynamic way to analyze words and symbols with tooltips for deeper understanding.

Successfully developed a news category classification model using fine-tuned BERT which can accurately classify any news text into its respective category i.e. Politics, Business, Technology and Entertainment.

Successfully developed a resume classification model which can accurately classify the resume of any person into its corresponding job with a tremendously high accuracy of more than 99%.

💥Fast State-of-the-Art Tokenizers optimized for Research and Production

Successfully fine-tuned a pretrained DistilBERT transformer model that can classify social media text data into one of 4 cyberbullying labels i.e. ethnicity/race, gender/sexual, religion and not cyberbullying with a remarkable accuracy of 99%.

Successfully established a text summarization model using Seq2Seq modeling with Luong Attention, which can give a short and concise summary of the global news headlines.

Successfully established a multiclass text classification model by fine-tuning pretrained DistilBERT transformer model to classify several distinct types of mental health statuses such as anxiety, stress, personality disorder, etc. with an accuracy of 77%.

Successfully developed a chatbot model which can provide accurate and concise responses to a wide variety of customer queries regarding the services offered by a particular company as well as general topics.

Successfully developed a text classification model to predict whether a given news text is fake or not by fine-tuning a pretrained BERT transformed model imported from Hugging Face.

Successfully developed a fine-tuned BERT transformer model which can accurately classify symptoms to their corresponding diseases upto an accuracy of 89%.

Successfully developed a fine-tuned DistilBERT transformer model which can accurately predict the overall sentiment of a piece of financial news up to an accuracy of nearly 81.5%.

The project aimed to push image captioning technology forward by combining recent advances in image recognition and language modeling to generate novel, descriptive captions that go beyond just naming objects and actions

Extract text content from an HTML page, process it, and extract unique words from the processed text. This notebook utilizes various text processing techniques including cleaning, normalization, tokenization, lemmatization or stemming, and stop words removal.