José Donato’s Post

Generative AI does not mean the “end of programming” People have been predicting the 'end of programming' since the dawn of time. It's one of those tired old tropes that refuses to die. In the mid-1990s it was all about CASE tools, closely followed by 4GLs. Then 'low code' tools were going to empower 'citizen developers' to build applications without engineers. Now it's the turn of generative AI and large language models (LLMs) to provide fuel for this lazy old cliché. The general idea is that generative AI will form a new abstraction that greatly simplifies building applications, transforming the role of an engineer. However, there is a world of difference between creating productivity tools and replacing the role of software engineer. Generative AIs are hard to work with as they are not deterministic or even that predictable, forgetting things, giving inconsistent responses, occasionally hallucinating, and being easily duped. https://ift.tt/C0gKRD1 #news #cto #tech

Learn how to train the YOLOv11 model using your custom dataset in Google Colab. From dataset labeling to importing, I'll guide you through the process, ensuring a seamless learning experience. Join me as I showcase YOLOv11's real-time object detection capabilities by demonstrating its versatility and accuracy in detecting various custom objects.  Access the full tutorial here: https://lnkd.in/d-iVjfuH Joel Nadar Harpreet Sahota 🥑 Timothy Goebel Ultralytics Muhammad Rizwan Munawar Dragos Stan Fiifi Amoah Shah Faisal Muhammad Faisal Piotr Skalski Anisha Udayakumar Aygun A. #yolov11 #computervision #yolov11custommodel

In this video, I have abstractly explained how you can use YOLOv5 for image recognition projects based on your dataset. Thank you for hitting the like 👍 button.

On-Device LLM with Voice! Here is a demo of LLM-based voice chatbot that runs locally on Raspberry Pi. It was inspired and based on Nick Bild's local_llm_assistant project, with some enhancements and modifications. The push button was taken from an old AIY Voice Kit. Luckily the old 7-in display has built-in speakers, so I don't have to get a separate speaker. Code: https://lnkd.in/gdzxyeHN

Take the step towards mastering algorithms with Venky.io! On April 12th, 2024, our comprehensive course will begin. Get ready to explore the world of algorithms and learn how to master them better than ever. Discover how to solve problems efficiently and master algorithms with us."

🚗 Automatic Number Plate Recognition (ANPR) Project 🚗 Excited to share my recent project where I developed an ANPR system using YOLO and PaddleOCR to detect and read vehicle license plates from video footage! Key Features: 🔹 Real-time Detection: Trained a YOLO model for fast and accurate license plate detection. 🔹 Text Recognition: Integrated PaddleOCR to extract license plate numbers with high accuracy. 🔹 Video processing: Used OpenCV for video processing and visualization. Check out the full project and code on GitHub: https://lnkd.in/drfUZewF Looking forward to your thoughts and feedback! #ANPR #MachineLearning #DeepLearning #ComputerVision #YOLO #PaddleOCR #OpenCV

Quickstart to Lightning Fabric - Lightning AI Lightning Fabric is a lightweight, production-ready framework that handles the heavy lifting of SOTA distributed training features, enabling researchers and machine learning engineers to train PyTorch models at scale. With just one Core API class - Fabric - users can configure their environment with arguments like devices, accelerator, strategy, and perform collective operations like broadcasting, gathering, and reducing. Fabric also has methods for setting up models and data loaders, loading and saving checkpoints, performing gradient clipping, and logging metrics. Users can launch their script or function to multiple processes with fabric.launch(), set up their data loader with fabric.setup_dataloaders(), and instantiate and set up their model and optimizer with fabric.setup(). The actual training loop is handled by fabric.backward(loss). Fabric provides maximum flexibility and control over the training logic, making it easy to adopt and add into existing PyTorch training loops without any need to restructure code. https://ift.tt/046CfGi #news #cto #tech

Retrieval Augmented Generation and Split Now we run it with our current setup, since all of our defaults are the values that existed before, we run it and we’ll get the same response.  Let’s play around with the feature flags to see if we get a different response. If we turn GPTModel to on – let’s see if we get a different response from GPT-4: (make sure to save the flag change!) Six hundred dollars was money enough to make half a dozen boys rich. Cool – now let’s turn the embeddingsModel flag on as well and see if that changes anything to use the more expensive embeddings model: Six hundred dollars was money enough to make half a dozen boys rich. Seems like for this question the embeddings model didn’t have any effect. Let’s change the question and see – maybe having a straightforward question doesn’t get strongly affected by these things. If we change this line: query = “Who is Joe Harper?” We get this answer: Joe Harper is a character mentioned in the context who seems to be a friend or acquaintance of Tom. He is described as being similarly dressed and equipped as Tom and is addressed by someone asking if he has seen Tom that morning. Joe Harper appears in response to these inquiries. Now let’s turn embeddingsModel off and see if there is a difference here. And we do get a different answer: Joe Harper is a character who is associated with Tom in the given context. He is described as being dressed and armed similarly to Tom, indicating that he is likely a close companion or friend of Tom. Additionally, there appears to be mention of Joe Harper’s mother, suggesting that Joe Harper is also important enough to be referenced in dreams by someone else. https://ift.tt/21UQCbs #news #cto #tech

Tagging David Cattermole, this is what happens when you feed OpenCV's feature tracker into #facebookresearch 's https://lnkd.in/dX7-2r8G kinda interesting to work with an old mate, gluing related technology together. What is really interesting to see if how wobbly these features are once you look at subpixel detail. #cotracker2 #trackingsolutions #rainbows 🌈 #badcodecs

Learn how to train the YOLOv11 model using your custom dataset in Google Colab. From dataset labeling to importing, I'll guide you through the process, ensuring a seamless learning experience. Join me as I showcase YOLOv11's real-time object detection capabilities by demonstrating its versatility and accuracy in detecting various custom objects.  Access the full tutorial here: https://lnkd.in/dXpnj7pb #yolov11 #computervision #yolov11custommodel