--- title: Best Towards AI posts — December 2024 | daily.dev description: The most upvoted Towards AI posts from December 2024, curated by the daily.dev community. canonical: https://daily.dev/sources/tai/best-of/2024/12 twitter:card: summary_large_image twitter:site: @dailydotdev og:url: https://daily.dev/sources/tai/best-of/2024/12 og:type: website og:site_name: daily.dev og:title: Best Towards AI posts — December 2024 | daily.dev og:description: The most upvoted Towards AI posts from December 2024, curated by the daily.dev community. og:image: https://media.daily.dev/image/upload/s--VAY5ToZt--/f_auto/v1724209435/public/daily.dev%20-%20open%20graph --- # Best of Towards AI — December 2024 1. 1 [](https://daily.dev/posts/build-your-llm-engineer-portfolio-a-3-month-roadmap-fwuhsbgly "Build Your LLM Engineer Portfolio: A 3-Month Roadmap") Article Towards AI · 2y Build Your LLM Engineer Portfolio: A 3-Month Roadmap A step-by-step guide to designing, refining, and showcasing a portfolio tailored for aspiring LLM engineers. Highlights the importance of practical, hands-on projects to stand out in the competitive AI job market and offers insights from the author's experience in crafting sophisticated GenAI applications and designing comprehensive product solutions. 107 2. 2 [](https://daily.dev/posts/10-no-nonsense-machine-learning-tips-for-beginners-using-real-world-datasets--kyc0rayup "10 No-Nonsense Machine Learning Tips for Beginners (Using Real-World Datasets)") Article Towards AI · 2y 10 No-Nonsense Machine Learning Tips for Beginners (Using Real-World Datasets) Get practical with machine learning by starting with simple models like Linear Regression and Decision Trees using real-world datasets from the UCI Machine Learning Repository. Focus on hands-on experimentation to build a strong foundation before diving into more complex models like neural networks. 71 3. 3 [](https://daily.dev/posts/computer-vision-object-detection-task-xelptwl1t "Computer Vision — Object Detection Task") Article Towards AI · 2y Computer Vision — Object Detection Task Object detection is an advanced version of object localization, involving identifying multiple objects and drawing bounding boxes around them. There are two types of models: two-stage models, which are outdated, and single-stage models, which are faster and easier to train. To solve the issue of predicting a fixed number of bounding boxes irrespective of actual objects, researchers developed techniques such as the Hungarian Matching Algorithm and various versions of the YOLO model. The post discusses the progression and implementation of these methods. 35 2 4. 4 [](https://daily.dev/posts/llama-explained--xzjw9sggf "LLaMA explained !") Article Towards AI · 2y LLaMA explained ! LLaMA, an advanced open-source large language model by Meta, brings several enhancements over conventional transformer architecture, such as grouped multi-query attention, RMS normalization, and rotary positional embeddings. These innovations result in more efficient computation and dynamic learning capabilities, making LLaMA a competitive choice for large language model applications. 34 5. 5 [](https://daily.dev/posts/llm-agents-and-agentic-design-patterns-smen7yfqk "LLM Agents and Agentic Design Patterns") Article Towards AI · 2y LLM Agents and Agentic Design Patterns Agentic AI is revolutionizing AI by enabling autonomous, distributed intelligent agents capable of real-time decision-making and dynamic problem-solving. A recent Berkeley course on LLM Agents offers deep insights into their history, reasoning patterns, and safety. Key highlights include the ReAct pattern for systematic exploration, importance of memory in LLMs, and the advantages of multi-agent systems for complex tasks. However, high costs and latency remain challenges. The field is rapidly evolving with new frameworks like Microsoft’s Autogen and Langchain's Langgraph. 33 6. 6 [](https://daily.dev/posts/how-to-deploy-ml-models-in-production-flawlessly--4jdjrw9t9 "How to Deploy ML Models in Production (Flawlessly)") Article Towards AI · 2y How to Deploy ML Models in Production (Flawlessly) When deploying machine learning models in production, it is crucial to focus on reliability, scalability, security, and maintainability. Using version control systems helps track different versions of your models, ensuring you can revert to stable versions if issues arise. The post offers insights into achieving reliable deployment for ML models in production environments. 30 7. 7 [](https://daily.dev/posts/llm-fine-tuning-guide-do-you-need-it-and-how-to-do-it-dyfcr5hqq "Llm Fine Tuning Guide: Do You Need It and How to Do It") Article Towards AI · 2y Llm Fine Tuning Guide: Do You Need It and How to Do It Fine-tuning a Large Language Model (LLM) is often unnecessary for many commercial applications, but it can be useful for tasks requiring specific chat formats, domain knowledge, or cost-effective, specialized tasks. Fine-tuning involves data preparation, including deduplication and removal of personal information, and can be done using techniques like LoRa (Low-Rank Adaptation) or QLoRA. Using reinforcement learning with human feedback (RLHF) or direct preference optimization (DPO) can align models with human preferences. For fine-tuning and hosting, cloud platforms like AWS SageMaker and collaborative tools like HuggingFace are recommended. 25 8. 8 [](https://daily.dev/posts/build-and-train-gpt-from-scratch-vyuumpjio "Build And Train GPT From Scratch") Article Towards AI · 2y Build And Train GPT From Scratch Learn how to build and train a Generative Pretrained Transformer (GPT) model from scratch using Python and PyTorch. Understand the internal mechanisms of GPT models, including self-attention and multi-head attention. Follow step-by-step instructions to construct the GPT architecture, tokenize data, implement self-attention, and train the model on a dataset. Discover techniques to improve model performance and optimize training and inference processes. 17 9. 9 [](https://daily.dev/posts/real-time-object-detection-using-yolov7-on-google-colab-qbj25eaqq "Real-Time Object Detection using YoloV7 on Google Colab") Article Towards AI · 2y Real-Time Object Detection using YoloV7 on Google Colab Learn how to perform real-time object detection using YOLOv7 on Google Colab in this detailed tutorial. Understand the structure of the training data and the bounding box representation used in YOLO models, and follow steps to apply the model to your videos. 15 10. 10 [](https://daily.dev/posts/how-to-build-a-proactive-agent-with-real-time-event-processing-eie6khhqg "How to Build a Proactive Agent with Real-Time Event Processing") Article Towards AI · 2y How to Build a Proactive Agent with Real-Time Event Processing Learn how combining streaming databases with large language models (LLMs) can enable proactive agents that take actions based on real-time events. The process involves setting up event listeners that can trigger the LLM to act upon specific conditions, much like setting an alarm. Utilization of SQL and materialized views in streaming databases allows efficient handling and monitoring of events, thus enabling the proactive behavior of intelligent agents. 12 11. 11 [](https://daily.dev/posts/llm-fine-tuning-guide-do-you-need-it-and-how-to-do-it-mjkrr3pjm "Llm Fine Tuning Guide: Do You Need It and How to Do It") Article Towards AI · 2y Llm Fine Tuning Guide: Do You Need It and How to Do It Fine-tuning language models (LLMs) is often unnecessary for many commercial applications due to the capabilities of modern LLMs. Instead, explore prompt engineering first. Fine-tuning becomes essential only for specific needs, such as unique chat formats or domain-specific knowledge. The process involves preparing quality instruction datasets, which can be sourced from open datasets or generated synthetically. Techniques like LoRa and QLoRa offer efficient alternatives to full model re-training. Lastly, reinforcement learning with human feedback and direct preference optimization can help align models with human preferences. 11 1 12. 12 [](https://daily.dev/posts/building-multimodal-rag-application-5-multimodal-retrieval-from-vector-stores-phmj7nao5 "Building Multimodal RAG Application #5: Multimodal Retrieval from Vector Stores") Article Towards AI · 2y Building Multimodal RAG Application #5: Multimodal Retrieval from Vector Stores Multimodal RAG combines textual and visual data to improve the retrieval process, enhancing the accuracy and detail of large language models. This guide covers setting up multimodal retrieval using the LanceDB vector database, highlighting installation, configuration, and ingestion of text and image data using LangChain. It concludes with a practical walkthrough for performing efficient multimodal searches. 11 13. 13 [](https://daily.dev/posts/master-a-b-testing-the-smart-way-stand-out-in-data-interviews-d5v3mf2tx "Master A/B Testing the Smart Way: Stand Out in Data Interviews") Article Towards AI · 2y Master A/B Testing the Smart Way: Stand Out in Data Interviews A/B testing is a vital tool in data-driven product development and is frequently discussed in data science interviews. This structured guide explains A/B testing through an example involving Spotify's recommendation algorithm, covering steps like understanding the business problem, identifying key metrics, formulating hypotheses, setting experiment parameters, designing the experiment, running and validating the test, interpreting results, and making launch decisions. By following these steps, organizations can confidently innovate and enhance user experiences. 10 [See all Towards AI archives](/sources/tai/best-of) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@graph":[{"@type":"CollectionPage","@id":"https://daily.dev/sources/tai/best-of/2024/12#page","url":"https://daily.dev/sources/tai/best-of/2024/12","name":"Best Towards AI Posts — December 2024","description":"The most upvoted Towards AI posts from December 2024, curated by the daily.dev community.","isPartOf":{"@type":"WebSite","url":"https://daily.dev"}},{"@type":"ItemList","@id":"https://daily.dev/sources/tai/best-of/2024/12#items","numberOfItems":13,"itemListElement":[{"@type":"ListItem","position":1,"url":"https://daily.dev/posts/build-your-llm-engineer-portfolio-a-3-month-roadmap-fwuhsbgly","name":"Build Your LLM Engineer Portfolio: A 3-Month Roadmap"},{"@type":"ListItem","position":2,"url":"https://daily.dev/posts/10-no-nonsense-machine-learning-tips-for-beginners-using-real-world-datasets--kyc0rayup","name":"10 No-Nonsense Machine Learning Tips for Beginners (Using Real-World Datasets)"},{"@type":"ListItem","position":3,"url":"https://daily.dev/posts/computer-vision-object-detection-task-xelptwl1t","name":"Computer Vision — Object Detection Task"},{"@type":"ListItem","position":4,"url":"https://daily.dev/posts/llama-explained--xzjw9sggf","name":"LLaMA explained !"},{"@type":"ListItem","position":5,"url":"https://daily.dev/posts/llm-agents-and-agentic-design-patterns-smen7yfqk","name":"LLM Agents and Agentic Design Patterns"},{"@type":"ListItem","position":6,"url":"https://daily.dev/posts/how-to-deploy-ml-models-in-production-flawlessly--4jdjrw9t9","name":"How to Deploy ML Models in Production (Flawlessly)"},{"@type":"ListItem","position":7,"url":"https://daily.dev/posts/llm-fine-tuning-guide-do-you-need-it-and-how-to-do-it-dyfcr5hqq","name":"Llm Fine Tuning Guide: Do You Need It and How to Do It"},{"@type":"ListItem","position":8,"url":"https://daily.dev/posts/build-and-train-gpt-from-scratch-vyuumpjio","name":"Build And Train GPT From Scratch"},{"@type":"ListItem","position":9,"url":"https://daily.dev/posts/real-time-object-detection-using-yolov7-on-google-colab-qbj25eaqq","name":"Real-Time Object Detection using YoloV7 on Google Colab"},{"@type":"ListItem","position":10,"url":"https://daily.dev/posts/how-to-build-a-proactive-agent-with-real-time-event-processing-eie6khhqg","name":"How to Build a Proactive Agent with Real-Time Event Processing"},{"@type":"ListItem","position":11,"url":"https://daily.dev/posts/llm-fine-tuning-guide-do-you-need-it-and-how-to-do-it-mjkrr3pjm","name":"Llm Fine Tuning Guide: Do You Need It and How to Do It"},{"@type":"ListItem","position":12,"url":"https://daily.dev/posts/building-multimodal-rag-application-5-multimodal-retrieval-from-vector-stores-phmj7nao5","name":"Building Multimodal RAG Application #5: Multimodal Retrieval from Vector Stores"},{"@type":"ListItem","position":13,"url":"https://daily.dev/posts/master-a-b-testing-the-smart-way-stand-out-in-data-interviews-d5v3mf2tx","name":"Master A/B Testing the Smart Way: Stand Out in Data Interviews"}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Sources","item":"https://daily.dev/sources"},{"@type":"ListItem","position":3,"name":"Towards AI","item":"https://daily.dev/sources/tai"},{"@type":"ListItem","position":4,"name":"Best of","item":"https://daily.dev/sources/tai/best-of"},{"@type":"ListItem","position":5,"name":"December 2024"}]}]} ```