MidCamp’s Post

How to get an LLM that is as good as GPT4 running locally in just a few minutes! Check this blog for detailed instructions: https://lnkd.in/g5X55utd

There are tons of courses on GenAI and LLMs. But, I was super impressed with the hands on course https://lnkd.in/gN_FbKei Highly recommended, since he gets into latest techniques like merge. He has already put a few models in HF using merge which are in leaderboard. So you are learning from a expert. #llms #genai #generatieveai #capgemini #llm #opensourceai   #leaderboard #learningeveryday #learningandgrowing #course2024

Ahh the patterns of production LLM applications… 🤖 First GPT 🏗️ First LLM App 🗃️ First RAG App 🕴️ First Agent App ⚖️ First Fine-Tuning 🔗 LangChain RAG 🦙 LlamaIndex RAG 📈 RAG Evaluation 🚀 Open-Source Production RAG 🔄 Agentic Production RAG 🌐 Domain-Adapted RAG with Fine-Tuning Get a handle on these concepts and code, start building, shipping, and sharing like a legend, and you’ll be creating real value for yourself, your company, and the world in no time! 🏗️ Start building for free, async: https://bit.ly/3UNimrg 🧑🤝🧑 Supercharge your AI Engineering journey with live cohort instruction, like-minded peers to meet you where you are and to show you where you’re going, accountability deadlines, and the most up-to-date curriculum you can find in the industry at the open-source edge: https://bit.ly/47GESVt #llms #production

How can Large Language Models (LLMs) enhance the onboarding process? BIOVIA #PipelinePilot's user-friendly interface leverages #AI-driven workflows and advanced analytics, enabling scientists at all career stages to utilize data effectively without needing programming expertise. However, getting started with any new tool can be time-consuming. In our latest blog by Dr. Mark Spillman, we demonstrate how on-demand expert assistance using LLMs can streamline the process of Pipeline Pilot protocol design and execution. Learn more about this integration here: http://go.3ds.com/OPq.

i know: Elon bad. But look at this: "An early version of Grok-2 has been tested on the LMSYS leaderboard under the name "sus-column-r." At the time of this blog post, it is outperforming both Claude 3.5 Sonnet and GPT-4-Turbo."

GPT5 unlocks LLM System2 Thinking? For human, problems often occur when we try to solve System 2 level complex problem, with System 1 fast, intuitive answer However this is exact the state of LLM now, there is no difference between answer 1+1 VS Complex math equation; All it does is just predict the best next words with fast, intuitive thinking; Fine tuning LLM = Human practicing something 10,000 hours, it produce better System 1 Intuition after intensive training for specific subject. ----- But how does human do the System 2 thinking? 1. We exchange speed for accuracy, we slow down, break down problems into steps, and play out possibilities to get the best answer; 2. Smartly switch between System 1 & System 2, our brain is an adaptive system, where it can save energy for simple problem, and reserve for complex ones Those 2 points are exactly what has been mentioned by @sama in an interview with @BillGates ---- So obviously we can expect GPT5 has much stronger reasoning ability + Adaptivity to switch between System 1 & System 2; But what can we do now to enforce System 2 thinking? There are 2 main methods: 1. Prompt engineering 2. Communicative agents In the video, I dived more details into current method with prompt engineering + agents setup: https://lnkd.in/gPBkDkpu

𝐎𝐩𝐞𝐧 𝐀𝐈 𝐢𝐬 𝐫𝐞𝐦𝐨𝐯𝐢𝐧𝐠 𝐂𝐡𝐚𝐭 𝐆𝐏𝐓 𝐏𝐥𝐮𝐠𝐢𝐧𝐬. 😳 𝐀𝐧𝐝 𝐡𝐞𝐫𝐞 𝐢𝐬 𝐰𝐡𝐚𝐭 𝐲𝐨𝐮 𝐧𝐞𝐞𝐝 𝐭𝐨 𝐤𝐧𝐨𝐰: 𝐏𝐡𝐚𝐬𝐞-𝐎𝐮𝐭 𝐨𝐟 𝐏𝐥𝐮𝐠𝐢𝐧 𝐁𝐞𝐭𝐚: The plugin beta journey concludes as we usher in the GPT era, with new conversations ceasing on March 19, 2024, and existing dialogues wrapping up by April 9, 2024. 𝐆𝐏𝐓 𝐒𝐭𝐨𝐫𝐞: 𝐀 𝐔𝐧𝐢𝐯𝐞𝐫𝐬𝐞 𝐨𝐟 𝐏𝐨𝐬𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐢𝐞𝐬: Transitioning from a vibrant ecosystem of over 1,000 plugins, the GPT Store now boasts an expansive collection of GPTs spanning writing, productivity, programming, education, and beyond, ready to redefine your experience and efficiency. 𝐀 𝐍𝐞𝐰 𝐂𝐡𝐚𝐩𝐭𝐞𝐫 𝐢𝐧 𝐀𝐈: This pivotal shift underscores our commitment to continuous improvement and innovation, ensuring the GPTs not only meet but exceed the functionalities previously offered by plugins, embodying the essence of full feature parity alongside groundbreaking new features. 𝐆𝐮𝐢𝐝𝐚𝐧𝐜𝐞 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐅𝐮𝐭𝐮𝐫𝐞: For those curious about the transition or eager to embark on creating their own GPT, comprehensive documentation and technical support await, paving the way for developers and enthusiasts alike to explore and contribute to this transformative journey. As we embrace this transformative phase, we invite you to join us in exploring the vast horizons of GPTs, where every feedback fuels our collective advancement. Together, let's step into the future, where the boundaries of AI are constantly expanded, and the potential for innovation knows no limits. 𝑨𝒓𝒆 𝒚𝒐𝒖 𝒓𝒆𝒂𝒅𝒚 𝒕𝒐 𝒓𝒊𝒅𝒆 𝒕𝒉𝒆 𝒘𝒂𝒗𝒆 𝒐𝒇 𝑺𝒎𝒂𝒓𝒕 𝑴𝒂𝒏𝒖𝒇𝒂𝒄𝒕𝒖𝒓𝒊𝒏𝒈? 𝑽𝒊𝒔𝒊𝒕 👉 𝒈𝒊𝒆𝒎𝒍𝒍𝒄.𝒄𝒐𝒎 𝒐𝒓 𝑫𝑴 𝒖𝒔 "𝑨𝑰𝟐𝟎𝟐𝟒" Credit: Greyson Zeng Oscar Carrasquillo #Fractionalleadership #AIautomation #Emergingtech #smartmanufacturing #Interimexecs

🚀 Day 35 of my 100-day challenge! Today's focus: Continued exploring recursion, mastering the quick sort algorithm using recursive techniques 🔄 #100DaysOfCode #Recursion #QuickSort #Algorithms #ProblemSolving

New Command R+ model. The open-source novelty already shows decent results on a level with GPT-4 and Claude 3. It is endowed with 104 billion parameters and a maximum of 128k contexts. Command R+ can write code, solve problems, support long dialogs with the user, and much more. Huggingface https://lnkd.in/gw6KwJr6 Try https://lnkd.in/gy79543e #tools #AIInnovation #NextGenModels #CuttingEdgeTechnology #NaturalLanguageProcessing #ArtificialIntelligence

The experience of fine-tuning a Large Language Model (LLM) having billions of parameters 🧠 💻 using just 20 carefully selected examples, and achieving flawless outputs that the base model could never replicate regardless of extensive 'prompt engineering', just leaves me in awe. #AIRevolution #MachineLearningMagic #InnovationInTech #DataScience #ArtificialIntelligence #TechTransformation