Research & Education

machine_learning_examples is an open-source GitHub repository offering a comprehensive collection of machine learning examples and tutorials. It covers a wide range of topics, including linear regression, logistic regression, neural networks, natural language processing, reinforcement learning, and more, with implementations in Python using libraries like NumPy, TensorFlow, and PyTorch. The repository serves as a valuable resource for students and practitioners looking to understand and apply machine learning concepts, often accompanying courses available on deeplearningcourses.com. Users are encouraged to clone the repository for easy updates rather than forking, as the content is frequently refreshed.

Natural-Language-Processing-Specialization is a GitHub repository containing all coursework, assignments, and slides for the Natural Language Processing Specialization offered by deeplearning.ai on Coursera. This resource is designed to equip individuals with state-of-the-art deep learning techniques necessary to build cutting-edge NLP systems. It covers topics such as sentiment analysis, language translation, text summarization, and chatbot development. The specialization is ideal for machine learning and AI students, as well as software engineers seeking a deeper understanding of NLP models and their application. Learners should have a working knowledge of machine learning, intermediate Python (including experience with a deep learning framework like TensorFlow or Keras), and proficiency in calculus, linear algebra, and statistics.

Tbilisi AI Meetups serves as a community platform dedicated to organizing and promoting gatherings centered around Artificial Intelligence, Machine Learning, and Data Science. The initiative's core mission is to facilitate knowledge exchange and networking opportunities for professionals and enthusiasts within the local AI community. By hosting regular meetups, it provides a space for individuals to discuss emerging trends, share insights, and collaborate on projects, thereby strengthening the regional AI ecosystem. The platform is designed to be accessible to anyone interested in these fields, from beginners to experienced practitioners, fostering a vibrant and collaborative environment.

Speechnotes is an AI-powered speech-to-text service designed for fast, accurate, and secure transcription and voice dictation. Users can dictate notes directly into an online notepad for free, or upload audio and video files for automatic transcription. The service supports various file types and languages, featuring speaker diarization, timestamping, captioning, and AI summaries. Speechnotes also offers a Chrome extension for voice typing across the web, an API for integration, and Zapier automation. With a focus on privacy, it ensures no human intervention in transcription and deletes recordings after processing. It also provides complementary tools like TTSReader for text-to-speech and Speechlogger for live captioning.

PageLM is a community-driven, open-source AI education platform designed to convert study materials into engaging and interactive learning experiences. Inspired by NotebookLM, it leverages state-of-the-art LLMs and TTS systems to create quizzes, flashcards, structured notes, and even podcasts from uploaded documents like PDFs, DOCX, Markdown, and TXT files. The platform offers features such as contextual chat for document interaction, automatic Cornell-style note generation, interactive quizzes with hints and scoring, and AI-powered podcast creation for on-the-go learning. It also includes a homework planner, exam simulator, AI debate partner, and a personalized study companion. PageLM supports multiple AI models including Google Gemini, OpenAI GPT, and Anthropic Claude, and is built with a modern tech stack including Node.js, React, and Docker.

nlp_overview is an open-source project dedicated to providing an up-to-date learning resource on modern deep learning techniques in natural language processing (NLP). It delves into the theoretical foundations and implementation specifics of deep learning models such as recurrent neural networks (RNNs), convolutional neural networks (CNNs), and reinforcement learning. The resource covers a wide array of NLP tasks and applications, including machine translation, question answering, and dialogue systems, and summarizes state-of-the-art results. The project aims to be a collaborative and open resource, guiding researchers and enthusiasts through emerging concepts, benchmark datasets, and code releases in the NLP field.

New-Bing-Anywhere is a versatile browser extension designed to enable users to access Bing's GPT-4 capabilities across a wide range of browsers, including Chrome, Firefox, Edge, Brave, Opera, Vivaldi, Arc, 360, and Yandex. This tool goes beyond simply enabling Bing outside of Edge; it integrates Bing's natural search and AI recommendations directly into search engine sidebars. This means that a single search can leverage both Google and Bing, aiming to provide more efficient and useful results. The extension also optimizes access for users in mainland China and Russia, supports multi-language interfaces, and offers features like quick switching between Bing and Google, and New Bing Image Create support. It is an open-source project, maintained by community support and donations.

PSEUDO.AI is an innovative AI-powered platform designed to effortlessly convert complex source code into clear, human-readable pseudocode. By leveraging OpenAI's trained GPT models, it bridges the communication gap between developers, designers, and stakeholders, streamlining collaboration and enhancing understanding of code. The tool supports a wide range of popular programming languages, including Java, Python, C++, and JavaScript. As a web-based platform, PSEUDO.AI requires no installation or downloads, making it easily accessible directly from any web browser. It is suitable for developers of all skill levels, from beginners seeking to understand complex code structures to experienced professionals looking to enhance productivity and clarity in their projects. PSEUDO.AI aims to transform complexity into clarity, freeing up time and energy for innovation.

The python-machine-learning-book-3rd-edition repository hosts the code examples for the third edition of the "Python Machine Learning" book. This resource is designed to complement the book, offering practical implementations of machine learning concepts using Python. It includes code for various chapters covering topics such as training algorithms, Scikit-Learn classifiers, data pre-processing, dimensionality reduction, model evaluation, ensemble learning, sentiment analysis, web application embedding, regression analysis, clustering, neural networks (from scratch and with TensorFlow), deep convolutional neural networks, recurrent neural networks, generative adversarial networks, and reinforcement learning. Users can explore these examples to deepen their understanding of the theoretical concepts presented in the book.

PyTorchNLPBook is a comprehensive companion repository for the book "Natural Language Processing with PyTorch," published by O'Reilly Media. It offers a rich collection of code and data designed to help users understand and implement NLP solutions using the PyTorch framework. The repository covers fundamental concepts such as PyTorch basics, foundational neural network components, and various NLP techniques including feed-forward networks, word embeddings, sequence modeling, and advanced topics like attention mechanisms and neural machine translation. It's an invaluable resource for anyone looking to learn and apply deep learning to natural language processing, providing hands-on examples and practical implementations directly from the book.

Women Who Do Data (W2D2) is a non-profit organization focused on empowering diversity within the artificial intelligence and data science fields. As a member-led initiative, W2D2 aims to advance women and underrepresented professionals in data, machine learning, and AI. The organization provides crucial opportunities for individuals to succeed in these rapidly evolving fields, promoting equitable progress and economic empowerment. By championing diverse talent, W2D2 strives to foster the creation of more inclusive and effective AI solutions, contributing to a more representative and innovative tech landscape.

PyPOTS (pronounced "Pie Pots") is a comprehensive Python toolkit designed for machine and deep learning on partially-observed time series data. It addresses the common issue of missing values in real-world time series by offering a wide array of state-of-the-art neural network models for tasks such as imputation, classification, clustering, forecasting, and anomaly detection. The library is built to simplify complex data analysis, allowing engineers and researchers to focus on core problems rather than data preprocessing. PyPOTS integrates with an ecosystem of tools like TSDB for dataset loading, PyGrinder for simulating missing data patterns, and BenchPOTS for standardized performance evaluation. It also supports hyperparameter optimization for all neural network models, making it a robust solution for scientific analysis of incomplete industrial and irregularly-sampled multivariate time series.

Reader by Jina AI is a powerful tool designed to optimize web content for Large Language Models (LLMs). It offers two primary functions: 'Read' and 'Search'. The 'Read' function converts any given URL into an LLM-friendly format, making it easier for agents and RAG systems to process and generate improved outputs. This includes the ability to read arbitrary PDF files from any URL and even generate captions for images that lack alt tags. The 'Search' function allows LLMs to access current world knowledge by searching the web for a given query and returning top results in an LLM-friendly format. It automatically fetches content from the top search results, bypassing issues related to browser rendering, JavaScript, and CSS. The tool supports various control options via request headers, including proxy settings, cache tolerance, and specific element targeting, making it highly adaptable for diverse use cases.

TheBigPromptLibrary is an extensive collection of prompts, system prompts, and LLM instructions designed for various large language model providers and solutions, including ChatGPT, Microsoft Copilot, Claude, Gab.ai, and Gemini. This resource offers significant educational value for individuals looking to improve their prompt writing abilities and understand the intricacies of creating custom GPTs. It also highlights the risks associated with prompt injection security. The library is organized into categories such as Articles, Tools and scripts, Custom Instructions, System Prompts, Jailbreak Prompts, and Instructions protections, making it a valuable learning resource for both beginners and experienced prompt engineers.

Amazon Q is a new generative AI-powered assistant from AWS, specifically designed to enhance productivity and streamline tasks within a business environment. It allows users to have conversations, solve problems, generate content, gain insights, and take action by connecting to a company’s information repositories, code, data, and enterprise systems. Amazon Q provides immediate, relevant information and advice to employees, accelerating decision-making and problem-solving while fostering creativity. It offers user-based plans and adapts interactions based on existing identities, roles, and permissions. AWS ensures customer content from Amazon Q is never used to train underlying models, maintaining data security and privacy. The tool supports over 40 built-in connectors for popular data sources like Amazon S3, Google Drive, Microsoft SharePoint, Salesforce, and Slack, and can be tailored to specific business needs.

AI Video Search Engine (AVSE) is an innovative tool designed to enhance video content interaction. Users can input questions to search across videos and receive relevant answers directly from the video content. Beyond simple search, AVSE offers the unique capability to chat with any video, transforming passive viewing into an interactive experience. This feature is particularly useful for extracting specific information, summarizing content, or engaging with educational and informational videos more deeply. The tool aims to save time and improve productivity by providing precise, context-aware results from video sources, making it easier to manage and discover content efficiently.

The Udemy-prompt-engineering-course GitHub repository serves as a comprehensive companion for the Udemy Prompt Engineering Course. It is built around Jupyter notebooks, small example applications, prompt files, datasets, screenshots, and diagrams, offering a hands-on approach to learning prompt engineering. The repository covers a wide array of topics including OpenAI API workflows, advanced prompting techniques, retrieval, embeddings, RAG, agent design and orchestration, LangChain, LangGraph, evaluation, vision, and image generation. It is particularly strong in coding-heavy sections, providing practical examples for developers and teams looking for reference notebooks for LLM application development.

UniWorld is an open-source project from PKU-YuanGroup focused on advancing visual AI through high-resolution semantic encoders. It provides a unified framework for understanding, generation, and editing of visual content. The project includes UniWorld-OSP2.0 for VLM-enhanced image-to-video generation, UniWorld-V2 for image editing using diffusion models, and UniWorld-V1 for a broad range of visual tasks. All data, models, training code, and evaluation code are open-sourced, making it a valuable resource for AI researchers and computer vision engineers. The framework demonstrates excellent performance across various tasks, including subject consistency, background consistency, and aesthetic quality in video generation, and precise instruction execution in image editing.

The TensorFlow-Machine-Learning-Cookbook is a comprehensive code repository published by Packt, designed to accompany the TensorFlow Machine Learning Cookbook. It offers all the necessary project files to work through the book, enabling users to gain practical experience with TensorFlow. The resource covers fundamental concepts such as variables, matrices, and data sources, progressing to advanced topics like Linear Regression, neural networks, Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Natural Language Processing (NLP). Each chapter's code is organized into folders, making it easy to follow along. It is compatible with Python 3 and requires libraries like TensorFlow, Numpy, Scikit-Learn, Requests, and Jupyter, running on Mac, Windows, and Linux without special hardware. This repository is ideal for those looking to deepen their understanding and application of Google's machine learning library.

Tensorflow-speech-recognition is an open-source project designed for speech recognition using Google's TensorFlow deep learning framework and sequence-to-sequence neural networks. It was developed as a replacement for caffe-speech-recognition. While the project is no longer actively maintained or up-to-date with the latest TensorFlow versions or state-of-the-art theory, it remains valuable for educational purposes. The repository provides various scripts for tasks like number classification, speaker classification, and speech-to-text, along with installation instructions for dependencies like pyaudio and portaudio. Users interested in modern speech recognition are advised to explore alternatives like Mozilla DeepSpeech or Whisper.

Beam ATG, also known as Beam, offers an advanced credit decisioning platform leveraging open finance methodologies and artificial intelligence. The tool is designed to streamline loan applications, clarify eligibility, and enhance the overall lending experience, particularly for African communities. Key features include the ability to increase conversion rates, boost loan volume, and significantly decrease administrative costs through automation and machine learning. Beam also focuses on mitigating bad debt risk with advanced algorithms and data analytics, promoting responsible lending. It provides powerful and easy-to-use APIs to unify consistent data and automate credit analysis, navigating the complex African credit landscape.

SiteScripter AI is a powerful Chrome extension designed to transform your web experience through intelligent automation and smart features. It seamlessly integrates into your browser, offering effortless configuration and intuitive commands. The tool provides smart, context-aware automation that adapts to your workflow, handling tasks like form filling, content summarization, and engaging in conversations with webpages. SiteScripter AI also empowers users to create compelling social media posts and streamline professional tasks with AI-generated email replies and job proposals. With a focus on efficiency, it aims to save users hours every week by automating repetitive online activities and providing quick insights from complex web content.

The OWASP Top 10 for Large Language Model Applications is a foundational resource for understanding and mitigating security risks in applications leveraging large language models. Housed under the OWASP GenAI Security Project, this initiative provides a broad consensus on the most critical security vulnerabilities specific to LLM technologies. It aims to offer practical, actionable, and concise security guidance for developers, data scientists, and security experts involved in designing and building LLM-powered applications and plugins. The project explores how conventional vulnerabilities manifest uniquely in LLM contexts and how traditional remediation strategies must adapt. It serves as a starting point for newcomers and a reference for experienced professionals in the evolving field of LLM application security.

Diglot is an AI-powered bilingual writing application designed specifically for non-native English speakers. It provides a comprehensive workspace where users can draft, translate, paraphrase, and refine their English writing without switching between multiple tools. Key features include an AI Writing Assistant for drafting and polishing, an AI Translator for in-context translation and comparison, a Grammar Checker for fixing errors, and a Paraphrasing Tool for improving fluency and originality. Additionally, it offers a Plagiarism Checker to ensure content originality before submission and writing templates for various document types like essays, emails, and reports. Diglot aims to streamline the writing process, making it easier to produce clear, fluent, and professional English content.