[Pytorch] Generative retrieval model using semantic IDs from "Recommender Systems with Generative Retrieval"

Python machine learning applications in image processing, recommender system, matrix completion, netflix problem and algorithm implementations including Co-clustering, Funk SVD, SVD++, Non-negative Matrix Factorization, Koren Neighborhood Model, Koren Integrated Model, Dawid-Skene, Platt-Burges, Expectation Maximization, Factor Analysis, ISTA, FISTA, ADMM, Gaussian Mixture Model, OPTICS, DBSCAN, Random Forest, Decision Tree, Support Vector Machine, Independent Component Analysis, Latent Semantic Indexing, Principal Component Analysis, Singular Value Decomposition, K Nearest Neighbors, K Means, Naïve Bayes Mixture Model, Gaussian Discriminant Analysis, Newton Method, Coordinate Descent, Gradient Descent, Elastic Net Regression, Ridge Regression, Lasso Regression, Least Squares, Logistic Regression, Linear Regression

The code to accompany the freeCodeCamp tutorial explaining how to use large language models to build a semantic book recommender.

Semantic IDs: How to train an LLM-Recommender Hybrid with steerability and reasoning on recommendations.

Bridging the Gap Between Semantic and Interaction Similarity in Recommender Systems

KDD2020 Improving Conversational Recommender Systems via Knowledge Graph based Semantic Fusion

GenRec: Generative Recommender Systems with RQ-VAE semantic IDs, Transformer-based retrieval, and LLM integration. Built on PyTorch with distributed training support.

Deep learning application to identify ingredients from cooking dishes images and recommend dishes to cook, given a set of ingredients. This application leverages NLP and Computer Vision to learn semantic knowledge using joint embeddings.

PHPMailer PHPMailer – A full-featured email creation and transfer class for PHP Test status codecov.io Latest Stable Version Total Downloads License API Docs Features Probably the world's most popular code for sending email from PHP! Used by many open-source projects: WordPress, Drupal, 1CRM, SugarCRM, Yii, Joomla! and many more Integrated SMTP support – send without a local mail server Send emails with multiple To, CC, BCC and Reply-to addresses Multipart/alternative emails for mail clients that do not read HTML email Add attachments, including inline Support for UTF-8 content and 8bit, base64, binary, and quoted-printable encodings SMTP authentication with LOGIN, PLAIN, CRAM-MD5, and XOAUTH2 mechanisms over SMTPS and SMTP+STARTTLS transports Validates email addresses automatically Protects against header injection attacks Error messages in over 50 languages! DKIM and S/MIME signing support Compatible with PHP 5.5 and later, including PHP 8.1 Namespaced to prevent name clashes Much more! Why you might need it Many PHP developers need to send email from their code. The only PHP function that supports this directly is mail(). However, it does not provide any assistance for making use of popular features such as encryption, authentication, HTML messages, and attachments. Formatting email correctly is surprisingly difficult. There are myriad overlapping (and conflicting) standards, requiring tight adherence to horribly complicated formatting and encoding rules – the vast majority of code that you'll find online that uses the mail() function directly is just plain wrong, if not unsafe! The PHP mail() function usually sends via a local mail server, typically fronted by a sendmail binary on Linux, BSD, and macOS platforms, however, Windows usually doesn't include a local mail server; PHPMailer's integrated SMTP client allows email sending on all platforms without needing a local mail server. Be aware though, that the mail() function should be avoided when possible; it's both faster and safer to use SMTP to localhost. Please don't be tempted to do it yourself – if you don't use PHPMailer, there are many other excellent libraries that you should look at before rolling your own. Try SwiftMailer , Laminas/Mail, ZetaComponents etc. License This software is distributed under the LGPL 2.1 license, along with the GPL Cooperation Commitment. Please read LICENSE for information on the software availability and distribution. Installation & loading PHPMailer is available on Packagist (using semantic versioning), and installation via Composer is the recommended way to install PHPMailer. Just add this line to your composer.json file: "phpmailer/phpmailer": "^6.5" or run composer require phpmailer/phpmailer Note that the vendor folder and the vendor/autoload.php script are generated by Composer; they are not part of PHPMailer. If you want to use the Gmail XOAUTH2 authentication class, you will also need to add a dependency on the league/oauth2-client package in your composer.json. Alternatively, if you're not using Composer, you can download PHPMailer as a zip file, (note that docs and examples are not included in the zip file), then copy the contents of the PHPMailer folder into one of the include_path directories specified in your PHP configuration and load each class file manually: <?php use PHPMailer\PHPMailer\PHPMailer; use PHPMailer\PHPMailer\Exception; require 'path/to/PHPMailer/src/Exception.php'; require 'path/to/PHPMailer/src/PHPMailer.php'; require 'path/to/PHPMailer/src/SMTP.php'; If you're not using the SMTP class explicitly (you're probably not), you don't need a use line for the SMTP class. Even if you're not using exceptions, you do still need to load the Exception class as it is used internally. Legacy versions PHPMailer 5.2 (which is compatible with PHP 5.0 — 7.0) is no longer supported, even for security updates. You will find the latest version of 5.2 in the 5.2-stable branch. If you're using PHP 5.5 or later (which you should be), switch to the 6.x releases. Upgrading from 5.2 The biggest changes are that source files are now in the src/ folder, and PHPMailer now declares the namespace PHPMailer\PHPMailer. This has several important effects – read the upgrade guide for more details. Minimal installation While installing the entire package manually or with Composer is simple, convenient, and reliable, you may want to include only vital files in your project. At the very least you will need src/PHPMailer.php. If you're using SMTP, you'll need src/SMTP.php, and if you're using POP-before SMTP (very unlikely!), you'll need src/POP3.php. You can skip the language folder if you're not showing errors to users and can make do with English-only errors. If you're using XOAUTH2 you will need src/OAuth.php as well as the Composer dependencies for the services you wish to authenticate with. Really, it's much easier to use Composer! A Simple Example <?php //Import PHPMailer classes into the global namespace //These must be at the top of your script, not inside a function use PHPMailer\PHPMailer\PHPMailer; use PHPMailer\PHPMailer\SMTP; use PHPMailer\PHPMailer\Exception; //Load Composer's autoloader require 'vendor/autoload.php'; //Create an instance; passing `true` enables exceptions $mail = new PHPMailer(true); try { //Server settings $mail->SMTPDebug = SMTP::DEBUG_SERVER; //Enable verbose debug output $mail->isSMTP(); //Send using SMTP $mail->Host = 'smtp.example.com'; //Set the SMTP server to send through $mail->SMTPAuth = true; //Enable SMTP authentication $mail->Username = 'user@example.com'; //SMTP username $mail->Password = 'secret'; //SMTP password $mail->SMTPSecure = PHPMailer::ENCRYPTION_SMTPS; //Enable implicit TLS encryption $mail->Port = 465; //TCP port to connect to; use 587 if you have set `SMTPSecure = PHPMailer::ENCRYPTION_STARTTLS` //Recipients $mail->setFrom('from@example.com', 'Mailer'); $mail->addAddress('joe@example.net', 'Joe User'); //Add a recipient $mail->addAddress('ellen@example.com'); //Name is optional $mail->addReplyTo('info@example.com', 'Information'); $mail->addCC('cc@example.com'); $mail->addBCC('bcc@example.com'); //Attachments $mail->addAttachment('/var/tmp/file.tar.gz'); //Add attachments $mail->addAttachment('/tmp/image.jpg', 'new.jpg'); //Optional name //Content $mail->isHTML(true); //Set email format to HTML $mail->Subject = 'Here is the subject'; $mail->Body = 'This is the HTML message body <b>in bold!</b>'; $mail->AltBody = 'This is the body in plain text for non-HTML mail clients'; $mail->send(); echo 'Message has been sent'; } catch (Exception $e) { echo "Message could not be sent. Mailer Error: {$mail->ErrorInfo}"; } You'll find plenty to play with in the examples folder, which covers many common scenarios including sending through gmail, building contact forms, sending to mailing lists, and more. If you are re-using the instance (e.g. when sending to a mailing list), you may need to clear the recipient list to avoid sending duplicate messages. See the mailing list example for further guidance. That's it. You should now be ready to use PHPMailer! Localization PHPMailer defaults to English, but in the language folder you'll find many translations for PHPMailer error messages that you may encounter. Their filenames contain ISO 639-1 language code for the translations, for example fr for French. To specify a language, you need to tell PHPMailer which one to use, like this: //To load the French version $mail->setLanguage('fr', '/optional/path/to/language/directory/'); We welcome corrections and new languages – if you're looking for corrections, run the PHPMailerLangTest.php script in the tests folder and it will show any missing translations. Documentation Start reading at the GitHub wiki. If you're having trouble, head for the troubleshooting guide as it's frequently updated. Examples of how to use PHPMailer for common scenarios can be found in the examples folder. If you're looking for a good starting point, we recommend you start with the Gmail example. To reduce PHPMailer's deployed code footprint, examples are not included if you load PHPMailer via Composer or via GitHub's zip file download, so you'll need to either clone the git repository or use the above links to get to the examples directly. Complete generated API documentation is available online. You can generate complete API-level documentation by running phpdoc in the top-level folder, and documentation will appear in the docs folder, though you'll need to have PHPDocumentor installed. You may find the unit tests a good reference for how to do various operations such as encryption. If the documentation doesn't cover what you need, search the many questions on Stack Overflow, and before you ask a question about "SMTP Error: Could not connect to SMTP host.", read the troubleshooting guide. Tests PHPMailer tests use PHPUnit 9, with a polyfill to let 9-style tests run on older PHPUnit and PHP versions. Test status If this isn't passing, is there something you can do to help? Security Please disclose any vulnerabilities found responsibly – report security issues to the maintainers privately. See SECURITY and PHPMailer's security advisories on GitHub. Contributing Please submit bug reports, suggestions and pull requests to the GitHub issue tracker. We're particularly interested in fixing edge-cases, expanding test coverage and updating translations. If you found a mistake in the docs, or want to add something, go ahead and amend the wiki – anyone can edit it. If you have git clones from prior to the move to the PHPMailer GitHub organisation, you'll need to update any remote URLs referencing the old GitHub location with a command like this from within your clone: git remote set-url upstream https://github.com/PHPMailer/PHPMailer.git Please don't use the SourceForge or Google Code projects any more; they are obsolete and no longer maintained. Sponsorship Development time and resources for PHPMailer are provided by Smartmessages.net, the world's only privacy-first email marketing system. Smartmessages.net privacy-first email marketing logo Donations are very welcome, whether in beer 🍺, T-shirts 👕, or cold, hard cash 💰. Sponsorship through GitHub is a simple and convenient way to say "thank you" to PHPMailer's maintainers and contributors – just click the "Sponsor" button on the project page. If your company uses PHPMailer, consider taking part in Tidelift's enterprise support programme. PHPMailer For Enterprise Available as part of the Tidelift Subscription. The maintainers of PHPMailer and thousands of other packages are working with Tidelift to deliver commercial support and maintenance for the open source packages you use to build your applications. Save time, reduce risk, and improve code health, while paying the maintainers of the exact packages you use. Learn more. Changelog See changelog. History PHPMailer was originally written in 2001 by Brent R. Matzelle as a SourceForge project. Marcus Bointon (coolbru on SF) and Andy Prevost (codeworxtech) took over the project in 2004. Became an Apache incubator project on Google Code in 2010, managed by Jim Jagielski. Marcus created his fork on GitHub in 2008. Jim and Marcus decide to join forces and use GitHub as the canonical and official repo for PHPMailer in 2013. PHPMailer moves to the PHPMailer organisation on GitHub in 2013. What's changed since moving from SourceForge? Official successor to the SourceForge and Google Code projects. Test suite. Continuous integration with Github Actions. Composer support. Public development. Additional languages and language strings. CRAM-MD5 authentication support. Preserves full repo history of authors, commits and branches from the original SourceForge project.

Implementation about a recommender System using RQ-VAE Semantic IDs

Official implementation of the paper "Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs"

Variable-Length Semantic IDs for Recommender Systems

Uses semantic search against a vector store to recommend similar products

基于HNSW的语义专家推荐系统

This is a repo listing some recommended papers on constituent parsing, dependency parsing and semantic role labeling on the CTB dataset.

# Frontend Mentor - Stats preview card component solution This is a solution to the [Stats preview card component challenge on Frontend Mentor](https://www.frontendmentor.io/challenges/stats-preview-card-component-8JqbgoU62). Frontend Mentor challenges help you improve your coding skills by building realistic projects. ## Table of contents - [Overview](#overview) - [The challenge](#the-challenge) - [Screenshot](#screenshot) - [Links](#links) - [My process](#my-process) - [Built with](#built-with) - [What I learned](#what-i-learned) - [Continued development](#continued-development) - [Useful resources](#useful-resources) - [Author](#author) - [Acknowledgments](#acknowledgments) **Note: Delete this note and update the table of contents based on what sections you keep.** ## Overview ### The challenge Users should be able to: - View the optimal layout depending on their device's screen size ### Screenshot  Add a screenshot of your solution. The easiest way to do this is to use Firefox to view your project, right-click the page and select "Take a Screenshot". You can choose either a full-height screenshot or a cropped one based on how long the page is. If it's very long, it might be best to crop it. Alternatively, you can use a tool like [FireShot](https://getfireshot.com/) to take the screenshot. FireShot has a free option, so you don't need to purchase it. Then crop/optimize/edit your image however you like, add it to your project, and update the file path in the image above. **Note: Delete this note and the paragraphs above when you add your screenshot. If you prefer not to add a screenshot, feel free to remove this entire section.** ### Links - Solution URL: [Add solution URL here](https://your-solution-url.com) - Live Site URL: [Add live site URL here](https://your-live-site-url.com) ## My process ### Built with - Semantic HTML5 markup - CSS custom properties - Flexbox - CSS Grid - Mobile-first workflow - [React](https://reactjs.org/) - JS library - [Next.js](https://nextjs.org/) - React framework - [Styled Components](https://styled-components.com/) - For styles **Note: These are just examples. Delete this note and replace the list above with your own choices** ### What I learned Use this section to recap over some of your major learnings while working through this project. Writing these out and providing code samples of areas you want to highlight is a great way to reinforce your own knowledge. To see how you can add code snippets, see below: ```html <h1>Some HTML code I'm proud of</h1> ``` ```css .proud-of-this-css { color: papayawhip; } ``` ```js const proudOfThisFunc = () => { console.log('🎉') } ``` If you want more help with writing markdown, we'd recommend checking out [The Markdown Guide](https://www.markdownguide.org/) to learn more. **Note: Delete this note and the content within this section and replace with your own learnings.** ### Continued development Use this section to outline areas that you want to continue focusing on in future projects. These could be concepts you're still not completely comfortable with or techniques you found useful that you want to refine and perfect. **Note: Delete this note and the content within this section and replace with your own plans for continued development.** ### Useful resources - [Example resource 1](https://www.example.com) - This helped me for XYZ reason. I really liked this pattern and will use it going forward. - [Example resource 2](https://www.example.com) - This is an amazing article which helped me finally understand XYZ. I'd recommend it to anyone still learning this concept. **Note: Delete this note and replace the list above with resources that helped you during the challenge. These could come in handy for anyone viewing your solution or for yourself when you look back on this project in the future.** ## Author - Website - [Add your name here](https://www.your-site.com) - Frontend Mentor - [@yourusername](https://www.frontendmentor.io/profile/yourusername) - Twitter - [@yourusername](https://www.twitter.com/yourusername) **Note: Delete this note and add/remove/edit lines above based on what links you'd like to share.** ## Acknowledgments This is where you can give a hat tip to anyone who helped you out on this project. Perhaps you worked in a team or got some inspiration from someone else's solution. This is the perfect place to give them some credit. **Note: Delete this note and edit this section's content as necessary. If you completed this challenge by yourself, feel free to delete this section entirely.**

# Frontend Mentor - Todo app solution This is a solution to the [Todo app challenge on Frontend Mentor](https://www.frontendmentor.io/challenges/todo-app-Su1_KokOW). Frontend Mentor challenges help you improve your coding skills by building realistic projects. ## Table of contents - [Overview](#overview) - [The challenge](#the-challenge) - [Screenshot](#screenshot) - [Links](#links) - [My process](#my-process) - [Built with](#built-with) - [What I learned](#what-i-learned) - [Continued development](#continued-development) - [Useful resources](#useful-resources) - [Author](#author) - [Acknowledgments](#acknowledgments) **Note: Delete this note and update the table of contents based on what sections you keep.** ## Overview ### The challenge Users should be able to: - View the optimal layout for the app depending on their device's screen size - See hover states for all interactive elements on the page - Add new todos to the list - Mark todos as complete - Delete todos from the list - Filter by all/active/complete todos - Clear all completed todos - Toggle light and dark mode - **Bonus**: Drag and drop to reorder items on the list ### Screenshot  Add a screenshot of your solution. The easiest way to do this is to use Firefox to view your project, right-click the page and select "Take a Screenshot". You can choose either a full-height screenshot or a cropped one based on how long the page is. If it's very long, it might be best to crop it. Alternatively, you can use a tool like [FireShot](https://getfireshot.com/) to take the screenshot. FireShot has a free option, so you don't need to purchase it. Then crop/optimize/edit your image however you like, add it to your project, and update the file path in the image above. **Note: Delete this note and the paragraphs above when you add your screenshot. If you prefer not to add a screenshot, feel free to remove this entire section.** ### Links - Solution URL: [Add solution URL here](https://your-solution-url.com) - Live Site URL: [Add live site URL here](https://your-live-site-url.com) ## My process ### Built with - Semantic HTML5 markup - CSS custom properties - Flexbox - CSS Grid - Mobile-first workflow - [React](https://reactjs.org/) - JS library - [Next.js](https://nextjs.org/) - React framework - [Styled Components](https://styled-components.com/) - For styles **Note: These are just examples. Delete this note and replace the list above with your own choices** ### What I learned Use this section to recap over some of your major learnings while working through this project. Writing these out and providing code samples of areas you want to highlight is a great way to reinforce your own knowledge. To see how you can add code snippets, see below: ```html <h1>Some HTML code I'm proud of</h1> ``` ```css .proud-of-this-css { color: papayawhip; } ``` ```js const proudOfThisFunc = () => { console.log('🎉') } ``` If you want more help with writing markdown, we'd recommend checking out [The Markdown Guide](https://www.markdownguide.org/) to learn more. **Note: Delete this note and the content within this section and replace with your own learnings.** ### Continued development Use this section to outline areas that you want to continue focusing on in future projects. These could be concepts you're still not completely comfortable with or techniques you found useful that you want to refine and perfect. **Note: Delete this note and the content within this section and replace with your own plans for continued development.** ### Useful resources - [Example resource 1](https://www.example.com) - This helped me for XYZ reason. I really liked this pattern and will use it going forward. - [Example resource 2](https://www.example.com) - This is an amazing article which helped me finally understand XYZ. I'd recommend it to anyone still learning this concept. **Note: Delete this note and replace the list above with resources that helped you during the challenge. These could come in handy for anyone viewing your solution or for yourself when you look back on this project in the future.** ## Author - Website - [Add your name here](https://www.your-site.com) - Frontend Mentor - [@yourusername](https://www.frontendmentor.io/profile/yourusername) - Twitter - [@yourusername](https://www.twitter.com/yourusername) **Note: Delete this note and add/remove/edit lines above based on what links you'd like to share.** ## Acknowledgments This is where you can give a hat tip to anyone who helped you out on this project. Perhaps you worked in a team or got some inspiration from someone else's solution. This is the perfect place to give them some credit. **Note: Delete this note and edit this section's content as necessary. If you completed this challenge by yourself, feel free to delete this section entirely.**

Easy understanding of the semantic segmentation using CNN with some recommended links.

Semantic Book Recommender (Python, LLM, OpenAI, LangChain, Gradio)

Recommended setup for releasing semantically using GitHub Actions workflows

Semantically Understanding Bias with LIME. Using the LIME-RS Algorithm to understand bias in recommender systems.

A recommender system based on semantic space clustering

Semantic book recommendation system using LLM embeddings, zero-shot classification, and emotion scoring with a Gradio dashboard.

This study aims to investigate the effectiveness of three Transformers (BERT, RoBERTa, XLNet) in handling data sparsity and cold start problems in the recommender system. We present a Transformer-based hybrid recommender system that predicts missing ratings and ex- tracts semantic embeddings from user reviews to mitigate the issues.

Daily arXiv paper recommender powered by semantic similarity to your Zotero library. Auto-analysis & Notion sync. 基于 Zotero 库语义相似度的每日 arXiv 论文推荐工具，AI 多模态分析 + 自动同步 Notion。

Official code for "Enriching Semantic Profiles into Knowledge Graph for Recommender Systems Using Large Language Models" (KDD'26)

No description available

Twitter semantic recommender based of entities detected by Wikipedia NER

A complex pharmaceutical recommendation system that makes use of cutting-edge technologies like LangChain for dynamic information retrieval, Neo4j for effective data storage, and SBERT for semantic embedding.

Sistema di raccomandazione film basato su conoscenza, che unisce MovieLens e DBpedia via SPARQL. Usa una strategia Multi-Armed Bandit con softmax per suggerire film semanticamente affini agli interessi dell’utente, considerando generi, tag, attori, registi e anni.