In this paper we present a multi-adapter retrieval augmented generation system (MARAGS) for Meta's Comprehensive RAG (CRAG) competition for KDD CUP 2024. CRAG is a question answering dataset contains 3 different subtasks aimed at realistic question and answering RAG related tasks, with a diverse set of question topics, question types, time dynamic answers, and questions featuring entities of varying popularity. Our system follows a standard setup for web based RAG, which uses processed web pages to provide context for an LLM to produce generations, while also querying API endpoints for additional information. MARAGS also utilizes multiple different adapters to solve the various requirements for these tasks with a standard cross-encoder model for ranking candidate passages relevant for answering the question. Our system achieved 2nd place for Task 1 as well as 3rd place on Task 2.

If some of the terminology I use here is unfamiliar, I encourage you to read my earlier article on LLMs first. There are teams that are employing ChatGPT or its competitors (Anthropic, Google's Flan T5 or PaLM, Meta's LLaMA, Cohere, AI21Labs, etc.) for real rather for cutesy demos. Unfortunately, informative content about how they are doing so is lost amidst marketing hype and technical jargon. Therefore, I see folks who are getting started with generative AI take approaches that experts in the field will tell you are not going to pan out. This article is my attempt at organizing this space and showing you what's working.

FastAPI has recently been making waves as an easy-to-use Python framework for creating APIs. If you're developing apps with FastAPI, you can add language processing capabilities to it by integrating Cohere's Large Language Models. In this article, you will learn how to create and finetune a Cohere sentiment analysis classification model, and generate predictions by making API calls to it using FastAPI. To follow this tutorial, you will need a Cohere account to generate an API key, create a finetuned model, and generate API calls. You also need a Python coding environment, such as VS Code.

Machine learning operations (MLOps) are a set of best practices for running machine learning models successfully in production environments. Data scientists and system administrators have expanding options for setting up their pipeline. However, while many tools exist for preparing data and training models, there is a lack of streamlined tooling for tasks like putting a model in production, maintaining the model, or monitoring performance. Enter vetiver, an open-source framework for the entire model lifecycle. Vetiver provides R and Python programmers with a fluid, unified way of working with machine learning models.

Machine learning operations (MLOps) are a set of best practices for running machine learning models successfully in production environments. Data scientists and system administrators have expanding options for setting up their pipeline. However, while many tools exist for preparing data and training models, there is a lack of streamlined tooling for tasks like putting a model in production, maintaining the model, or monitoring performance. Enter vetiver, an open-source framework for the entire model lifecycle. Vetiver provides R and Python programmers with a fluid, unified way of working with machine learning models.

It is hard to agree on the best tools to serve models in production because each problem is unique, and their solutions have different constraints. Therefore, I wanted to choose a solution or a set of tools that would benefit as many people as possible. The solution should be simple enough so that it takes only a few minutes to whip up a working prototype and serve it online and, if needed, can scale to larger-scale problems. The core component of this solution is the BentoML package. It is one of the latest promising players in the MLOps landscape and has already amassed half a million downloads on GitHub.

Content platforms thrive on suggesting related content to their users. The more relevant items the platform can provide, the longer the user will stay on the site, which often translates to increased ad revenue for the company. If you've ever visited a news website, online publication, or blogging platform, you've likely been exposed to a recommendation engine. Each of these takes input based on your reading history and then suggests more content you might like. As a simple solution, a platform might implement a tag-based recommendation engine -- you read a "Business" article, so here are five more articles tagged "Business."

Bot Libre now allows you to create generic deep learning analytics and train them through our web API. Deep learning analytics can be used for a wide array of purposes to analyze and make predications on data. This example shows how to train a deep learning analytic to play checkers. You can use either the Bot Libre deep learning library, or the TensorFlow deep learning library. You can choose the inputs, outputs, and layers.

Few years ago, I shared first machine learning story about insurance claim prediction. It's based on python code with logistic regression algorithm to build simple classification model as demonstration purpose. In 2020, it should be the year of Automatic Machine Learning (Auto ML) to make machine learning process clean, simple, fast and everyone can taste it, even peoples haven't knowledge in machine learning or data science. Recently, due to job related, I'm helping my customer to explore/evaluate data science and machine learning platform solution. That's surprise me that Azure Machine Learning (AML) is enhanced a lot and really provided an end-to-end solution platform and take care wide ranges of end users, from newbie to expert.

We click on our newly deployed model and will find see a list of attributes, of which the relevant one is the "Scoring URI". Remember talking about API endpoints? This Scoring URI is our endpoint for the model we have created, meaning we can send ("do a request") something to this endpoint (input) and will receive a response (output). And if we format the request correctly, we will get back what we want: a price per night estimate given our input columns. This part might get a bit confusing for those who have no experience working with HTTP requests, but just keep following the steps and you'll get it to work!