You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
I think you know, things are maybe a little less ideal than how you've painted them. First, on hitting a wall on compute. Many results have depended on a large investment in compute to be achievable: top pretrained LMs, stylegan, RL game bots etc. Unless something changes, andpic.twitter.com/CoHcIVwlUE
The something I expect to change is a focus on using the same hardware more effectively (optimization!). I think there's a lot of reasons to believe that there's a lot to be gained here: 1. Historically, software efficiency gains often dwarf anything you can get out of hardware
example- EfficientNet came out recently and showed just how inefficient previous vision networks were (and there's a lot of attention being paid to these!) We're talking order of magnitude here. And it does work (I use it!)pic.twitter.com/SOJ9iOJtWO
In my own work, I've been able to shrink down DeOldify 10x and get better performance with this mindset. There were simply just bad decisions made at the software level previously (mostly stuff I grabbed from elsewhere that was considered pretty much SOTA!).
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
