7/ ConvNets Match Vision Transformers – evaluates a performant ConvNet architecture pretrained on JFT-4B at scale; observes a log-log scaling law between the held out loss and compute budget.
INNOVATION
-

Matryoshka Diffusion Models for High-Resolution Image Video Synthesis
By
–
3/ Matryoshka Diffusion Models – introduces an end-to-end framework for high-resolution image and video synthesis; involves a diffusion process that denoises inputs at multiple resolutions jointly and uses a NestedUNet architecture…
-

Spectron: End-to-End Spoken Language Model Surpasses Existing Systems
By
–
4/ Spectron – a spoken language model trained end-to-end to directly process spectrograms; it can be fine-tuned to generate high-quality accurate spoken language; surpasses existing spoken language models in speaker preservation and semantic coherence.
-

Evaluate LLM Performance: Metrics and Benchmarking Strategies
By
–
Master LLMs: Top Strategies to Evaluate LLM Performance In this video, we look into how to evaluate and benchmark Large Language Models (LLMs) effectively. Learn about perplexity, other evaluation metrics, and curated benchmarks to compare LLM performance. With practical tools
-

Spooky Black Light SDXL Fine-Tune Model Released
By
–
Spooky black light SDXL fine tune anyone? https://
replicate.com/fofr/sdxl-blac
k-light
… -
Traffic Simulator Project Showcases MathWorks Career Growth
By
–
Explore this traffic simulator created by Manuel during his internship! 🚗🌟
— MathWorks (@MathWorks) 29 octobre 2023
Now a full-time MathWorker, read about his journey from Grad School to a career at MathWorks👉 https://t.co/r8NqM46z5R #MathWorksCareers pic.twitter.com/3g0e9V2louExplore this traffic simulator created by Manuel during his internship! Now a full-time MathWorker, read about his journey from Grad School to a career at MathWorks http://
spr.ly/6014u6Bbe #MathWorksCareers -
FAA conservative approach to novel aviation technology regulation
By
–
FAA is an extremely conservative regulator and generally takes the position that anything novel which cannot be proven to not interfere with flight systems is presumptively a threat to flight systems. This exists in some tension with other goals we have for flight infrastructure
-

DoReMi: Optimizing Data Mixtures for Faster Language Model Pretraining
By
–
Check out DoReMi: Optimizing Data Mixtures Speeds Up Language Model Pretraining, and some of the papers it cites. By @sangmichaelxie et al.


