Vespa.ai’s Post

Exciting development in optimization! 👏 Researchers from the University of Tokyo have introduced ADOPT, a new adaptive gradient method that addresses the convergence issues of Adam without the need for specific Hyperparameter tuning. ADOPT achieves an optimal convergence rate and shows superior performance across multiple tasks, including image classification and large language models. [The paper has been accepted in NeurIPS 2024] For anyone working with adaptive optimizers, this is a must-read! Check out the paper for detailed insights and theoretical analysis. Arxiv Link: https://lnkd.in/g4sZvDzd GitHub Implementation: https://lnkd.in/ga2NUTfj #AI #MachineLearning #DeepLearning #Optimization #Research

I highly recommend this course! As a PM, I can easily tell that we need to be prepared for using AI constantly in our lives in order to boost our productivity, check quality, correct possible failures. There are infinite uses.  Just finished the course “Build Your Own GPTs” by Alina Zhang! Check it out: https://lnkd.in/dYZc7AHH #chatbotdevelopment.

I just finished "Build your Ownn GPT's" course. The course quick and opens minds for new ideas in the limitless universe of the AI.  Thanks Alina Li Zhang  Check it out: https://lnkd.in/dFhgsP3b #chatbotdevelopment.

Finished the course “Build Your Own GPTs” ! Its amazing how you we can customize the AI response to a certain format. Surely, this can be incorporated for optimization in many business processes.

🚀 New technique unlocks significant performance gains for Large Language Models (LLMs) Just came across an intriguing paper by Matteo Pagliardini, Pierre Ablin, and David Grangier on the state-of-the-art AdEMAMix optimizer. Feeling inspired, I took my keyboard and implemented it myself using PyTorch. You can check out implementation here: https://lnkd.in/d9q2BN7Q During my initial experiments, I noticed improvements compared to the baseline optimizer AdamW. Stay tuned for a blog post where I explore deeper. You can find original paper in the comments section. #AI #MachineLearning #Optimization #PyTorch #LLM #LargeLanguageModels #DeepLearning #Innovation

Exciting times as we pivot to an era where English becomes the language for human-machine interactions! 🌟 Context and domain knowledge are key, and Alina Li Zhang's "Build Your Own GPTs" course is a game-changer. Perfect for leveraging AI in our evolving landscape!  #BuildYourOwnGPTs #TechInnovation #PwCLabs

OpenAI released a new o1 model today 🔥It is 5-6 (!) times better than the past strongest neuron GPT-4o.This is the very secret project that has been worked on for so long development.The new model shows itself 5 times better in math problems and 6 times better in writing code!This strong increase in quality is due to the fact that the model “thinks” before giving you the answer.Access starts today.Source | Artificial intelligence 🤖

Just finished "Build Your Own GPTs" by Alina Zhang! Check it out: https://lnkd.in/ggN47Nm2 #chatbotdevelopment. Exiting to learn more about the evolution of AI in project management

I can't wait to create a few GPTs 🤖 that could influence an untapped market! As I finished the course “Build Your Own GPTs” by Alina Li Zhang, I am reminded that Ai has capabilities that were only mere thoughts years ago. Now, we are capable of so much more, and I cannot wait to see what else we can accomplish. The course was a great introduction and I highly recommend you check it out!  https://lnkd.in/eSmdHjK3 #chatbotdevelopment.

Just learned to build my own GPT, finally, an AI that agrees that Plato’s ‘Allegory of the Cave’ is just another way to explain a bad market forecast. Now, if only my classmates could see the light!  https://lnkd.in/gfEiveVE #chatbotdevelopment.