The AI Scientist: Agentic, fully-automated research pipeline for under $15 per paper Researchers have just created an AI system that can conduct entire research projects from start to finish, potentially revolutionizing how scientific discoveries are made. This "AI
MACHINE LEARNING
-

VQ4DiT: Efficient Post-Training Vector Quantization for Diffusion Transformers
By
–
VQ4DiT Efficient Post-Training Vector Quantization for Diffusion Transformers paper page: https://
huggingface.co/papers/2408.17
131
… The Diffusion Transformers Models (DiTs) have transitioned the network architecture from traditional UNets to transformers, demonstrating exceptional capabilities -

New Hub feature displays carbon emissions from model training
By
–
New feature on the Hub! Carbon emissions emitted during training now show up on the model card!
(requires model authors to fill that info first) Hopes it will prompt more people to show the carbon emissions of their model training! Thanks a lot to the team who pushed -
Lab advances vision and robotic learning integration
By
–
Our lab’s new work that shows deeper integration of vision and robotic learning! 🤩🦾 https://t.co/MDORPjlkmh
— Fei-Fei Li (@drfeifei) 2 septembre 2024Our lab’s new work that shows deeper integration of vision and robotic learning!
-
Fine-tuning and Merging LLMs: Techniques and Best Practices
By
–
If you're interested in fine-tuning and merging LLMs, here's a well-edited video of a talk I gave in July. We talk about: – How to create a dataset
– SFT techniques
– How to merge models Thanks to @Tunehq_ai for the invitation! https://
youtu.be/8aFqLVQjHTw?si
=K8KpM55G_93wwZmM
… -
Responsible AI Must Keep Pace with ML Capabilities
By
–
@IValeraM pointing out how many open questions there are in causal and ethical machine learning . “We can’t have a technology whose capabilities advance fast without similar speed in advances in responsible ai”. #DLI2024 #Indaba2024 #xamxamle
-
ChatGPT-4o does the unimaginable and is the best
By
–
ChatGPT-4o can do what people can't even imagine. Its just the best
-
Vocal Ambiguity: Reverse Correlation Analysis Framework
By
–
Webpage: https://
rosielab.github.io/vocal_ambiguit
y/
…
Paper: https://
arxiv.org/abs/2406.05515
Code for reverse correlation: https://
github.com/neuro-team-fem
to/vocal_ambiguity
… Huge thanks to collaborators Jean Julien Aucouturier, Henny Yeung and Yue Wang! -

Free Fast.ai Machine Learning Course Part 2 Available Online
By
–
…and if you too want to take this course, it's available for free here: https://
course.fast.ai/Lessons/part2.
html
… -

Deep Learning Indaba 2024 Showcases Africa’s AI Talent and Startups
By
–
Get excited . We are rolling out the red carpet as we kickoff the gathering and the showcase of Africa’s AI expertise, communities, startups, and emerging talent at this year’s Deep Learning Indaba @DeepIndaba in Senegal @MahtarMbow #DLI2024 #Indaba2024 #xamxamle
