Machine Learning

Announcing Lightspeed’s latest investment in the AI space — Large language models (LLMs) have taken the world by storm in recent months, and for good reason — LLMs are uniquely suited to comprehend and extract insights from unstructured data, outperforming previous machine learning approaches that either didn’t work at all or were extremely brittle. 70% of enterprise data is…

In today’s recreational coding exercise, we learn a more advanced and robust Monte Carlo approach for model parameter fitting, which also allows us to calculate the Bayesian evidence of a model and perform model selection. …

A process of prompt engineering, software engineering, trial and error, and elbow grease — Unlike tabular data, datasets for computer vision tasks are unstructured — think gobs of pixels, heaps of labels, bags of tags, and some sometimes-structured metadata. Nevertheless, those of us working in computer vision still need to…

The NYU AI School returned with its fourth iteration this summer, May 31 — June 4, 2023. The free week-long course was launched and generously funded by the NYU ML² Machine Learning for Language Lab and organized by students from the CDS and NYU’s Courant Institute. It includes hands-on labs…

NC, SA, GPL, and other acronyms you don’t want to see in the open source license of the model you are using — A (biased) history of open sourcing AI libraries and models I joined the Google Brain team in 2015 right as TensorFlow was open sourced. Contrary to popular belief, TensorFlow was not the secret sauce behind Google’s success at that point in time. Only a handful of researchers had used it, and it took several years before it transformed Alphabet’s most…

Engaging with our own grief and comprehending loss is a profoundly intimate experience. This often all-consuming process can lead to feelings of alienation, compounded by the emotional labour of continually narrating our personal journeys to provide reassurance to others. Yet, when we subject these deeply personal and distressing experiences to…

A State-of-the-Art LLM Better than LLaMa for Free — The Falcon models are state-of-the-art LLMs. They even outperform Meta AI’s LlaMa on many tasks. Even though they are smaller than LlaMa, fine-tuning the Falcon models still requires top-notch GPUs with more than 40 GB of VRAM. Well, this was before the introduction of QLoRa. QLoRa quantizes LLMs in 4-bit…

By Shruti Datta Gupta, Security Researcher — Security teams commonly track, triage, and fix product security bugs or vulnerabilities using a ticketing system. However, as an organization grows, so does the security team, splitting into smaller teams that focus on specific areas of expertise. Often, these teams create and follow their own, disparate processes for categorizing and…