Written in collaboration with AI/ML GDE Aashi Dutt. Introduction In the preceding two articles, we successfully learned how to prepare your own dataset and fine-tune the Gemma model using supervised fine-tuning with LoRA. Just as a journey needs companions, an ML project needs its model deployed. It’s like peanut butter without jelly — incomplete. In the final part of this series, we’ll walk you through deploying the fine-tuned Gemma model on Android, enabling you to have a complete end-to-end application at your fingertips. Before we begin, ensure that you have Android Studio instal ..read more

Enhancing Web Image Accessibility for Visually Impaired Individuals with Gemini Pro Vision and Google Cloud Platform Problem The inability of visually impaired individuals access image information due to the lack of adherence to W3C web accessibility initiatives by websites. Currently, about 60% of websites lack meaningful alternate text for their images. Moreover, it is unfeasible to retroactively add descriptive text to all existing websites manually. Two mins short Project Introduction and our story https://medium.com/media/350efbd62c0a343dc4dec2353120f5a9/href Demo 1 https://medi ..read more

Introduction to coroutine behavior through playful examples Sometimes you think you know coroutines and then after a while, you’re like “wait, do I really know coroutines?” This series serves as spin off from Programming Android with Kotlin: Achieving Structured Concurrency with Coroutines intended to help strengthen everyday coroutine understanding through playful explorations. We [the authors] always had sincere intentions with the book: While [coroutine] concepts are important if you want to master coroutines, you don’t have to understand everything right now to get started and be ..read more

I’ve always had a soft spot for PEZ dispensers and have been collecting them for over 30 years. These quirky little collectibles come in an incredible variety of shapes and characters. But there’s more to PEZ than just fun; some dispensers can be quite valuable depending on their age and variation. To address the challenge of identification, I harnessed the power of AI to create an image classification model that could help identify these subtle differences. Use Case: Identifying PEZ dispensers Identifying the precise type of PEZ dispenser isn’t always easy. Take Mickey Mouse dispensers, for e ..read more

[Mar 2024] ML Community — Highlights and Achievements Let’s explore highlights and accomplishments of the vast Google Machine Learning communities over the month. We appreciate all the activities and commitment by the community members. Without further ado, here are the key highlights! Featured Stories ML Developer’s Journey AI/ML GDE Rubens Zimbres (Brazil) shared how he increased the number of readers and followers on his Medium channel. From Mar 2023 to Mar 2024 (1 year), readers have increased by 700% and followers increased by 800%! One of his articles, Augmenting Gemini-1.0-Pro with Kno ..read more

Have you ever stared at a photo of a delicious dish and wondered what it was or how to make it? With the power of AI and image recognition, that question can now be a thing of the past. In this article, we’ll explore how to leverage Gemini Vision Pro, a large language model from Google AI, alongside Colaboratory (Colab), a free Jupyter Notebook environment in the cloud, to generate recipes based on an image. What is Gemini Vision Pro? Gemini Vision Pro is a cutting-edge vision model from Google AI capable of understanding and interpreting visual content. It can analyze images and ext ..read more

Taking advantage of the Gemini's multi-modal input to create trailers for any videos. This year on February 15, Google announced the release of Gemini 1.5, this new version brought many improvements, and on top of impressive improvements in the language domain, this model can process a huge input context of up to 1 million tokens, to make it even better it is was trained as a multimodal model, this means that is can natively process text, images, audio or video. This combination of different input types and huge context got me excited with the opportunity to process long videos, so I rev ..read more

Screenshot of Google AI Studio with Gemini 1.5 Pro model selected, and my multi-modal prompt with my L.A. trip photos How accelerated has been the evolution of Generative AI technologies! People are impressed by Multi-modal LLMs, that can understand and generate text, images, videos and audio using a single end-to-end model. In this post, I demonstrate how to use a great recent multimodal LLM — Gemini 1.5 Pro — for the se case of generating a blog post solely from photos and videos taken on a trip. In the end, I also talk briefly about some popular multi-modal LLM architectures and public ..read more

Image Source: https://dataedo.com/asset/img/banners/blog/cartoons.png Google recently launched Gemini 1.5 Pro model, which is a mid-sized multimodal model optimised for scaling across wide range of tasks. In this blog, we will learn how Gemini 1.5 Pro model can help us during software development. This blog is an improved and recent version of my previous blog, How Generative AI improves the productivity of Software developers All examples in this blog use the freeform prompt in Google AI Studio and Gemini 1.5 pro model. Below are some of the ways in which Gemini 1.5 Pro can help sof ..read more

Mathematical word problem-solving has long been recognized as a complex task for small language models (SLMs). To reach a good level of performance with these models, researchers often train SLMs to generate Python code or by using ensembling techniques, associated with consensus or majority vote. The challenge here is to use Google’s Gemma model, with less than 2 billion parameters and with safeguards against generating code to solve these Grade School Math problems. Here I will use Microsoft’s Orca-Math dataset, a high quality synthetic dataset of 200K math problems obtained through a multi ..read more