Nikhil Sonti’s Post

Video Walkthrough: DBmarlin 4.2 with Co-pilot AI 🤖 See how DBmarlin's new Co-pilot AI works and how it offers tuning recommendations for optimising your database and SQL statements. #DBmarlin #ApplicationPerformance #Database #Blog #AI #CoPilot

Mistral AI just dropped new ‘mixture of experts’ model with a torrent link! Love their confidence 🙌 Mixtral 8x22B's capabilities are still unknown and the torrent leads to four files totalling 262GB... probably difficult to run locally 🙄 But I think we can only expect something super cool from this team! #mistral #mistralai #mixtral #ai #artificialintelligence #generativeai #genai #llms #llm #largelanguagemodels #deeplearning #data #datascience #aidevelopment #aiengineering #vcs #paris #france #europe

Bluescarf.ai has published an amazing step by step guide to finetune a LLM. Follow Bluescarf.ai to keep yourself updated with the newest AI news. Fine-Tuning the Llama 2 Model In this step-by-step guide, we delve into the fine-tuning process of the Llama 2 Model, showcasing how to enhance performance with cutting-edge techniques. 🔹 Step 1: Set up your environment with essential packages. 🔹 Step 2: Import critical libraries for seamless model training. 🔹 Step 3: Configure model parameters for optimal efficiency with QLoRA and 4-bit quantization. 🔹 Step 4: Load and fine-tune the model, pushing boundaries with advanced training setups. 🔹 Step 5: Save your finely-tuned model for deployment. 🔹 Step 6: Test the model’s prowess with real-world prompts to ensure peak performance. Join us on this journey to elevate your AI models to the next level! #ArtificialIntelligence #MachineLearning #Llama2 #ModelFineTuning

Tired of spending countless hours manually labeling your data? 😩 Auto-annotation is here to transform your workflow! 🚀 Tuba’s innovative platform leverages the power of Large Video Models (LVMs) to automate the annotation process. 📊✨ Experience the Benefits: Accelerate Your Projects ⚡ Enhance Accuracy 🎯 Focus on Insights 🔍 How it Works: Our platform utilizes advanced AI models 🤖 to automatically identify and label objects, actions, and scenes within your images and videos. This significantly speeds up the data preparation process without compromising quality. ⏱️📈 Ready to Get Started? Watch our tutorial to see auto-annotation in action! https://lnkd.in/efrSfK-6 🎥👀 Try it for free at Tuba.AI 🌟 Need a customized solution? Request a demo: https://lnkd.in/dhBGt9QS 💬 #AutoAnnotation #ComputerVision #AI #Efficiency #DataLabeling #Tuba #AIvision #AIsolutions

Presenting Qdrant support in Aurelio AI Semantic-Router! A library to build a decision-making layer for your AI agents! Rather than waiting for slow, unreliable LLM generations to make tool-use or safety decisions, Semantic-Router uses vector space magic to route our requests using semantic meaning. Qdrant is now available as an index provider to ingest and retrieve appropriate routes in your Semantic-Router-powered decision-making layer. Find more about getting started with Semantic-Router and Qdrant: 📚 Integration docs: https://buff.ly/4cpajY2 🎥 Semantic-Router video course: https://buff.ly/4cklGAm

Q&A with Text data is a fascinating Generative AI use case. See this demo in my YouTube channel to see its immense value and potential #generativeai #llm #openai

🚗 Excited to share my latest project on 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐜 𝐍𝐮𝐦𝐛𝐞𝐫 𝐏𝐥𝐚𝐭𝐞 𝐑𝐞𝐜𝐨𝐠𝐧𝐢𝐭𝐢𝐨𝐧 (𝐀𝐍𝐏𝐑)! Using YOLOv9 with the Glen-C model for object detection and 𝐎𝐂𝐑 for text recognition, I achieved impressive results with the following model summary: Model: YOLOv9 with Glen-C, 467 layers, 25.4M parameters, 102.5 GFLOPs Performance: Precision: 0.902 Recall: 1.00 mAP@50: 0.97 mAP@50-95: 0.71 The dataset was collected from Roboflow and annotated using the Segment Anything Model (SAM), which greatly streamlined the annotation process! Super thrilled with these results and looking forward to applying this in real-world use cases! 🎯🚀 Source Code : https://lnkd.in/dHpZnRJV #MachineLearning #ComputerVision #ANPR #OCR #YOLOv9 #GlenC #AI #DeepLearning #Roboflow #SAMModel #TechInnovation #Project

Netlify's CEO is based!! Mathias Biilmann live-coded an AI agent with Langbase x Netlify to bring back GeoCities. You can simulate any domain, so fun. Excited to open-source it today with Netlify. It's been used quite a lot already: 🔹 GPT-4o (you can try any LLM) 🔹373.24K tokens 🔹551 runs (API Reqs) 🔹0.93 seconds avg latency Netlify x Langbase. Check out the open pipe (live AI playground) below ❯ AI GeoSim agent pipe ↳ https://lnkd.in/gi-QNp9N ❯ Live Demo (hint: add /something and see it simulate that new path) ↳ https://lnkd.in/gw-wDrRY Blog post on how it works: ↳ https://lnkd.in/gnwtCJ8F You can also watch the talk's video here. Pretty interesting. ↳ https://lnkd.in/gc4y7Rmt Netlify just published a developer guide on it here: ↳ https://lnkd.in/gqSbqsVn And finally, here's a technical guide we wrote: Build and deploy this open-source AI agent yourself. A step-by-step guide: ↳ https://lnkd.in/ggfP6Dpx Let's go! 👊 #AI #Agents #GeoCities #Langbase #Netlify

Just finished the course “Agentic AI for Developers: Concepts and Application for Enterprises” by Kumaran Ponnambalam! Check it out: https://lnkd.in/eqgF_yzn #generativeai #aisoftwaredevelopment #aiagents. Good overview of agentic concepts including examples with code.

New PhpStorm Line Completion with Local AI in v2024.1 is amazingly accurate. It suggested this line JUST from the name of the migration file. Wow. Another reason to name things properly according to standards: then AI assistants will help you on your path.