Raju Shreewastava’s Post

What an incredible day it was at the "Build With AI" event hosted by Google Developer Group (GDG) Silicon Valley at the Google Cloud office, Sunnyvale! the event brought together enthusiasts and experts to dive into Gemini abilities and the latest in AI technologies. The day kicked off with an invigorating keynote by Asim Husain, setting a high bar for the sessions that followed. Vikram Tiwari gave an enlightening talk on Gemini Abilities and Responsible AI, which really opened my eyes to the ethical dimensions of technology. Peter Danenberg’s workshop on Function Calling with Gemini was a hands-on highlight, pushing all of us to explore and experiment. The sessions by Victor Dibia, PhD and the dynamic duo, Debi Cabrera and Chloe 🍤 Condon, were not only informative but also incredibly inspiring, showing practical applications of GenAI in developing multi-agent systems and transforming data into actionable insights. One of the more reflective moments came from the career talk by Mariane Bekker which focused on creating inclusive spaces in tech. And, of course, the cherry on top was winning not one, but two Google swags from the giveaway—definitely a lucky day! Is there such a thing as too much Google swag? Asking for a friend who definitely isn't me—because we all know, more is always better! 😆 A massive shoutout to Sakshi Gupta for her flawless hosting and to all the GDG volunteers who made the day run so smoothly. The connections made and lessons learned are invaluable. Here's to more such enriching experiences! Can’t wait for the next one! #BuildWithAI #GoogleDevelopersGroup #GenAI #Gemini #SiliconValley

I am 46 and I STILL participate in hackathons and coding contests. In 2022, my team finished in 2nd place in the finals of Google Cloud Next hackathon - https://lnkd.in/gWqCwyW2 We were very close but lost out on the top prize. This year, we are determined to go ONE STEP BETTER and I need your support! I strongly believe GenAI will change our lives for good - whether it's technology, education, medicine, social media, healthcare or any other domain, a BIG CHANGE is around the corner. As part of my ongoing interest to do a deep-dive on GenAI, I alongside my teammates Abhinav Jain and Abhay Jidge have submitted our innovative idea in the Google Cloud presents GenAI Hackathon. Our app is GenCreator - a video generation platform that aims to revolutionize content creation by transforming ideas into text prompts and then into engaging video content. Our solution is 100% native to Google Cloud - it leverages the power of Google Vertex AI, Gemini, and Google Cloud Stack to offer a seamless and intuitive interface for users across various industries, including education, marketing, and entertainment. You can see our submission video here - a detailed walkthrough of the problem statement and our proof of concept in action - https://lnkd.in/gajH9fz9 What is your take on our idea? Wish my team best of luck for the shortlist announcement. Every like/ comment/ share will help our submission in this keenly contested hackathon so I request your support! Thanks Hack2skill for this opportunity!   #GenAI #hack2skill #Hack2skillcontest #googlecloud #GenAIHackathon #GenerativeAI #VertexAI #Gemini #GoogleGemini #GCP #Hackathon #learning P.S - The shortlist gets announced tomorrow (April 1) and this is NOT an April Fool’s joke

"Thrilled to have participated in the Google AI Fest Campus organized by Google Developers! 🚀 Engaged in an inspiring hackathon and successfully tackled challenges, earning a certificate in the process. Grateful for the opportunity to delve into the world of AI alongside talented peers. Excited for what lies ahead in the realm of technology and innovation! #GoogleAI #Hackathon #AIEnthusiast"

🏎 #fastai is a #DeepLearning library that provides high-level components to beginners and practitioners to quickly get started with standard models, while also providing more seasoned researchers with low-level components that can be mixed and matched to build new approaches. 🤝 And did you know that fastai is fully integrated with Comet? 👇 Get started building your own projects with fast.ai + Comet in the quickstart Colab:

Late to the announcement, but I am thrilled to share that I'm a winner in Google Cloud Powered Code Vipassana Project Saadhna Cycle 2 Hackathon! 🎉 My project, 'Gemini/Vertex AI Powered Daily Quote, Poem, Cover Image posting/Generation Website,' was selected as a winners and has been listed in Top 5 . 🏆 Read more about Project Saadhna Cycle 2 winners on annnoucement article: https://lnkd.in/g8jHSeiu Read about it on my Google Cloud Medium article : https://lnkd.in/gQrSC5kj TL;DR : 𝐌𝐲 𝐩𝐫𝐨𝐣𝐞𝐜𝐭 𝐮𝐬𝐞𝐬 𝐚 𝐜𝐨𝐦𝐛𝐢𝐧𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐀𝐈 𝐭𝐨𝐨𝐥𝐬 𝐭𝐨 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐞 𝐚𝐧𝐝 𝐩𝐨𝐬𝐭 𝐛𝐥𝐨𝐠𝐬 𝐜𝐨𝐧𝐭𝐚𝐢𝐧𝐢𝐧𝐠 𝐝𝐚𝐢𝐥𝐲 𝐪𝐮𝐨𝐭𝐞𝐬 𝐰𝐢𝐭𝐡 𝐩𝐨𝐞𝐦𝐬 𝐚𝐧𝐝 𝐜𝐨𝐯𝐞𝐫 𝐢𝐦𝐚𝐠𝐞𝐬. 𝐈𝐭'𝐬 𝐚 𝐬𝐦𝐚𝐥𝐥 𝐞𝐱𝐚𝐦𝐩𝐥𝐞 𝐨𝐟 𝐡𝐨𝐰 𝐀𝐈 𝐬𝐞𝐫𝐯𝐢𝐜𝐞𝐬 𝐰𝐢𝐭𝐡 𝐬𝐨𝐦𝐞 𝐜𝐨𝐝𝐞 𝐜𝐚𝐧 𝐛𝐞 𝐛𝐫𝐨𝐮𝐠𝐡𝐭 𝐭𝐨𝐠𝐞𝐭𝐡𝐞𝐫 𝐟𝐨𝐫 𝐜𝐫𝐞𝐚𝐭𝐢𝐯𝐞 𝐜𝐨𝐧𝐭𝐞𝐧𝐭 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧. While it was a sprint effort <2weeks, it was a lot of fun. Some services I integrated includes Zenquotes, Quotables, Google Gemini, LimeWire, Stability AI, Google Imagen. I'm incredibly grateful to Code Vipassana Team and Abirami Sukumaran for this opportunity and excited to see my project in the leaderboard! Special mention, Mohit Bhimrajka who the lead assigned to me who has been very helpful throughout the process.🚀 I went through all finalist submissions and was impressed by the variety of ideas brought to life by fellow #CodeVipassana Peers. 👏 If you are also interested in joining the fun, Code Vipassana Season 7 starts in September - codevipassana.dev #imagen #limewire #gcp #arjunraghunandanan #aicontentgeneration #dailyquotes #machinelearning #generativeai 🎉#ProjectSaadhna

🔥 XGBoost, which stands for Extreme Gradient Boosting, is an #OpenSource, scalable, distributed gradient-boosted decision tree (GBDT) #MachineLearning library. Although XGBoost isn't new, it remains a popular favorite for supervised #ML tasks like regression, classification, and ranking. XGBoost algorithms are designed, in particular, to be highly efficient, flexible and portable. And XGBoost is seamlessly integrated with Comet! 👇 Try the Colab to get started: https://lnkd.in/exsuUReS #MLOps #DataScience #Integration

🚀 I'm excited to share that I recently participated in the Google AI Hackathon, an incredible event held at the Innovation and Digital Development Agency! Our team, the Neurolists, was among the 21 selected from over 250 teams to compete. Alongside my talented teammates—Javidan Ganbarov (AI Engineer), Atilla Khalilli (Full Stack Developer), and Zeynal Mardanli (Frontend Developer)—we tackled the challenge of Early Disease Detection and Prediction. Using TensorFlow we developed a sequential #deeplearning model to predict diseases based on patient data and symptoms. We then created a website to integrate our model, resulting in a robust platform that suggests potential diseases and recommends appropriate doctors. Although we did not place in the top three, the experience was invaluable. I want to extend my heartfelt thanks to my teammates for their hard work and dedication. 🙌 Congratulations to the winning teams! 🏆 A big thank you to the Innovation and Digital Development Agency, Tech Academy, Bakcell and all the other sponsors for organizing this fantastic event. It was an excellent opportunity to strengthen our skills, collaborate with fellow AI enthusiasts, and push the boundaries of innovation. 💡🤖 #ai #hackathon #google

🚀 Thrilled to Announce Our Victory at the Google AI Hackathon! 🚀 I’m beyond excited to share that I recently participated in the Google AI Hackathon , where I teamed up with incredible minds from around the globe. Together, we created SparkAI, a game-changing product that transforms user ideas into dynamic content across various platforms. Imagine this: you have a spark of an idea, and with SparkAI, that idea morphs into a fully-fledged article, engaging post, or captivating short video. The theme of the hackathon was to harness Google’s GEMINI and AI tools, combined with low-code platforms, to bring innovative concepts to life. We embraced this challenge head-on, using BuildShip for the backend and FlutterFlow for the frontend. It was a whirlwind of reading documentation and watching tutorials, but we thrived under pressure. I’m incredibly grateful for my phenomenal team: Sidhu Di, Dhyey Bhaiya, , , Godfred Awu and Hari Bhaiya. Special shoutout to and for their relentless dedication, staying online despite different time zones, and teaching me so much along the way. All our sleepless nights and hard work paid off—WE WON! 🏆 We clinched the "Most Impressive Low-Code Implementation" award, securing a $5000 cash prize and an exclusive opportunity to meet the Google Labs team for official recognition. This journey has been nothing short of electrifying, and I’m so proud of what we’ve achieved together. Here’s to many more innovations and victories! #Gemini #Innovation #LowCode #WinningTeam #ProudMoment #AI #Teamwork #TechVictory #Google #Hackathon #AIProjects #GlobalCompetition #InnovationRecognition #GoogleIO #ArtificialIntelligence #TeamSuccess #HackathonExperience #TechCommunity #FutureOfAI 🌐

I fine-tuned the Mistral-7B-Instruct-GPTQ model on Google Colab, targeting the finance-alpaca dataset to enhance its financial analysis capabilities.  Employing the Accelerate library enabled distributed training, while PEFT with QLoRA significantly streamlined the model's adaptability without bloating its size or computational demand.  The integration of a 4-bit quantization via GPTQConfig further optimized the model's efficiency, making it more resource-effective. Gradient checkpointing and LoRA's selective parameter updates were pivotal, ensuring an efficient training process with detailed progress logging and evaluation. The entire process is documented in the attached Colab notebook. #generatieveai #finetuning #mistral7b #alpaca #googlecolab #machinelearning https://lnkd.in/gAPdRMNU

🏎 fastai holds a special place in my heart as it is the library I used in my first Kaggle competitions. It provides high-level components to #DeepLearning beginners and practitioners to quickly get started with standard models, while also providing more seasoned researchers with low-level components that can be mixed and matched to build new approaches. And now fastai is fully integrated with Comet! 👇 Get started building your own projects with fast.ai + Comet in the quickstart Colab: