I'm curious, what are the differences between T5, Flan-T5, and Flan-UL2 for fine-tuning? Does the instruction tuning matter at all, once you're fine-tuning?
I'm curious about the prompt you used, if you're willing to share (I'm interested in blend words, which this is somewhat related, but not quite - these are blended in a semantic sense, not textual)
I'll take the opportunity to request better editing ergonomics: the ability to connect from Jupyter/notebook-supporting editors and IDEs; the ability to open/edit .ipynb files from the local disk and/or Github without having to first put them on Google Drive.
Colab/Jupyter and friends are reinventing many wheels around editing code, and it would be nicer for them to support tools like Jupytext.
It would be really nice if I didn't have to fill out a form to see a demo-- having used Scuba I'd love to advocate for Interana at my current company, but it keeps getting pushed down my TODO list because of the extra friction
"We have seen the video but have been unable to replicate the experience it shows," a Facebook representative told CNET in a statement. The company also once again detailed how its display-filtering technology, which surfaces news to user feeds, has built-in safeguards to keep potential spam out (emphasis mine):
"Newsfeed is an automated system that is designed to deliver the most relevant content to you and your friends. The technology evaluates hundreds of factors, including your relationship to the poster, the type of content, the click-through rate (where appropriate), and people hiding similar posts from their feed. In real time, it decides what to display to you and what to filter for both Top News and Most Recent. It also includes systems that attempt to identify and block spam. Links have a history of the most abuse and are given the most scrutiny. As a result of all of these factors, a given link may be shown or filtered to people differently at different times."
1. Is the source available anywhere? I'm curious to see how it works.
2. Is there a way to connect the structure displayed here, to the rendered version in the PDF? To visually display the subcomponents?