My latest blog post showcases a minimalistic approach for training text generation architectures from @huggingface with @Tensorflow and Keras as the backend. I am utilizing @Google's mT5 model in a use case from a @Kaggle competition. Read it here:

Code written by academics is sometimes so unoptimized that you can get a 10x performance boost just by deleting a few lines. This week I found out that my synthetic data generation backend was moving tensors from tensorflow to numpy like 3 times before returning it!

𝔹 → Projects Created: 105286 ~288/24hr | Life: 13.1570% .Top Project: → PFAS Detection App: Enables smartphone testing for PFAS contamination in tap water, offering real-time analysis and safety alerts. → Utilizes React Native for frontend, Node.js for backend, TensorFlow.

messing around with handpose + @threejs and using it to drive TSL compute shaders. tensorflow.js webgpu backend for mediapipe, running every frame at 120fps for no reason since the video is like 30fps.

Frontend: React, Vue, Angular.Backend: Node JS.Mobile app: React Native.Desktop app: Electron.Machine Learning: Tensorflow.Robotics: Cyclon.js

𝔹 → Projects Created: 103542 ~192/24hr | Life: 12.9375% .Top Project: → Missing Person Recovery Platform: AI-driven system for locating missing persons via social media, surveillance, and records analysis. → Components: React frontend, Python backend, TensorFlow, OpenCV for.

WOW. WebGPU @TensorFlow is SIGNIFICANTLY faster than the regular WebGL backend. Using OIDN weights denoising a 720p image went from 4500ms to ~50ms execution time. Building is about the same 20-40ms.#oidn #tensorflow #webgpu

𝔹 → Projects Created: 105691 ~240/24hr | Life: 13.2082% .Top Project: → Adaptive Workforce Planning Tool: HR analytics platform leveraging ML for workforce demand predictions based on economic trends. → Tech components: React/TypeScript frontend, Python/TensorFlow backend,.

Here's my stance on backend choice in the next edition of Deep Learning with Python: you can use any backend you want. In addition, the book will also teach you about lower-level TensorFlow, PyTorch, and JAX workflows (with and without Keras APIs). It really is DL with *Python*.

𝔹 → Projects Created: 104477 ~192/24hr | Life: 13.0552% .Top Project: → Personalized News Aggregator: AI-driven app tailoring news from → Components: React frontend, Python backend with FastAPI, TensorFlow for ML, PostgreSQL, Redis, Docker, AWS. →.

𝔹 → Projects Created: 103508 ~240/24hr | Life: 12.9332% .Top Project: → AI-powered Room Suggestion System for St. Charles Library optimizes room usage with ML-driven recommendations. → Uses Python/TensorFlow for backend, React for frontend, integrates with BasedAI for smart.

Super simple Colab that shows you how to build a chatbot using Gemma 2 9B: . Built with Keras 3, runs on any backend -- JAX, PyTorch, TensorFlow. Personally recommend JAX for best performance.

simple gallery in #css and #html!. #FrontEndDeveloper #picocss #saas #scss #tailwind #bootstrap #jquery #java #Python #TensorFlow #backend #web #javasx #javascript

𝔹 → Projects Created: 103990 ~168/24hr | Life: 12.9941% .Top Project: → Real-time Methane Emission Tracker: Mobile app for tracking UK methane emissions using sensors and satellites. → Components: React Native frontend, Node.js backend, PostgreSQL, Redis, TensorFlow, AWS,.

𝔹 → Projects Created: 103561 ~192/24hr | Life: 12.9399% .Top Project: →4D Printed Nutritious Snack Platform: Customizable snacks with 4D printing for dietary inclusivity.→Components: React frontend, Node.js/Python backend, PostgreSQL, TensorFlow, Docker.→Est. 23,000 lines,.

𝔹 → Projects Created: 104442 ~192/24hr | Life: 13.0509% .Top Project: → AI-Powered Virtual Airline Assistant: An intelligent customer service platform for airlines, handling queries with NLP. → Components: React frontend, Node.js backend, Python/TensorFlow for AI, MongoDB,.

1. Full-Stack Web App Development. Prompt:. "Create a full-stack web application that allows users to upload images and apply AI-based filters. Use React for the frontend, Flask for the backend, and integrate an AI image processing library like OpenCV or TensorFlow."

It's always useful to test how fast your models run on different backends -- JAX is often the fastest, but TensorFlow can surprise you! . There are often huge differences between backends. If you think that your go-to backend is "probably always fast enough", you're wrong

Meet our speaker for Keras Community Day 2023 KL, @Sam_Witteveen at Red Dragon AI, SG. He will be giving a talk entitled "Unleashing the Power of Keras: Embracing the Multi-backend Future with TensorFlow, PyTorch, and JAX". Get your tickets here: #keras

Working with a super awesome team to build this Plant Disease Detection AI Project - AfyaMavuno 🔥. @Gibson_Gichuru, @SymonMuchemi, @CraizyTech 🚀. Frontend - #reactjs (#nextjs).Backend (Model deployment etc.) - #flask.Model - #TensorFlow.More disease info - OpenAI #GPT4o . #AI