It is not uncommon for Keras 3 + JAX to be 3-5x faster than PyTorch models pulled from research repos on GitHub or from Hugging Face. Though the typical figure is more like 30% faster.
MACHINE LEARNING
-
KerasCV Segment Anything: Zero-shot Image Segmentation with Prompt Points
By
–
KerasCV now includes a Segment Anything implementation, which enables you to do image segmentation using only "prompt points". No training data needed. With the JAX backend, it runs ~5x faster on GPU than the original PyTorch implementation. Told you JAX was fast. Guide:
-

Top 10 Deep Learning Algorithms: CNNs to DBNs Explained
By
–
Unlock the potential of AI with this breakdown of the top 10 deep learning algorithms by @techprofreeteam! From CNNs to DBNs, the future is here. Elevate your tech knowledge by following @ingliguori for the latest in #DeepLearning and #AI! #TechTrends #MachineLearning
-
Data Augmentation via Embedding Cone Sampling with Vec2Text
By
–
this was a neat little read. the idea is to do data augmentation with embeddings. they randomly sample around an embedding and then decode with vec2text. there is a trick to randomly sampling while not leaving the embedding manifold; they try to sample within an embedding "cone"
-
Groq MLAgility Performance Benchmarks and Argonne Workshop Recording
By
–
Open-source enthusiast curious about our perf results? Check out MLAgility where we show our performance on hundreds of off-the-shelf models compared to other vendors. Recording from our recent @argonne_lcf workshop for researchers using #inference to accelerate their work!
-
Neural Operators Optimize Medical Catheter Design in Science Advances
By
–
Our work on using Neural Operators to design a medical catheter just got published in Science Advances. https://t.co/4HJAFIbQgt
— Prof. Anima Anandkumar (@AnimaAnandkumar) 4 janvier 2024
A catheter is a flexible tube to draw fluids out of a human body. Unfortunately, catheters frequently cause bacterial infections by bacteria swimming… pic.twitter.com/TwiysyswncOur work on using Neural Operators to design a medical catheter just got published in Science Advances. http://
tensorlab.cms.caltech.edu/users/anima/ca
theter.html
…
A catheter is a flexible tube to draw fluids out of a human body. Unfortunately, catheters frequently cause bacterial infections by bacteria swimming -
Model Optimization: Fixed 8fps Frame Rate Control Strategy
By
–
Aye, there’s no fps control at the moment, it’s fixed to 8fps, but you can choose how many frames to generate. I think there’s also lots of room for optimisation of this model to get it faster.
-

Call for Speakers: Share Your GenAI and ML Insights
By
–
Is one of your #NewYearsResolutions to speak at #DataAISummit? It’s time to make it a reality! Submit your proposal to share your #genAI, data science, and #ML insights. Deadline is this Friday: https://
bit.ly/3sOkYt0 -
Join AI Communities and Learn Machine Learning Together
By
–
I also put emphasis on joining communities and interacting with peers in machine learning for collective progress. Staying engaged is crucial for success as the field advances and to learn with others. All relevant links in the guide. Start learning AI: https://
github.com/louisfb01/star
t-machine-learning
… -
Flexible Machine Learning Mastery Guide with Free Resources
By
–
My guide provides a flexible path to mastering machine learning, catering to different learning preferences with free resources, from quick YouTube tutorials to thorough articles, books and courses.