Profile avatar
jsulz.com
I like pretty things, functional things, funny things, food things, and computer things. Used to do devops/WordPress things @lexblog.bsky.social and devex/cloud infra things at @pantheon.io Now helping make things go fast at 🤗 @hf.co
149 posts 217 followers 99 following
Regular Contributor
Active Commenter

Forest bathing. A common Washington state pastime.

I've been working on cutting down my, "I'm sorry, I meant to be more clear in my last prompt" messages too. There is a price to being polite. Good breakdown by Julien Delavande, @sashamtl.bsky.social, and Régis Pierrard.

Just get the Windows box back up and running to troubleshoot some user reports about slow downloads in WSL. It'll be easy. I *won't* spend the entire morning downloading software and rebooting the machine.

Learned GCP was out by seeing my dog door monitor was broken (backed by a cloud SQL instance). How am I going to replay/track the events of her going in and out the dog door during this outage? These are the important questions.

🎉 Big update! The #InternetArchive has launched a new version of GifCities, our search engine for vintage GeoCities GIFs. Search better. Blink more! Check it out ➡️ blog.archive.org/2025/06/09/k... 🧵

As a side-effect of the @gradio-hf.bsky.social hackathon, we may have just accidentally created the largest open source collection of MCP servers. If you are curious about what can be done with MCP you should really check them out, it’s pretty wild: huggingface.co/spaces?filte...

A philosophy to live by (also, check out those beards 🧔🧔)

"It started a few years ago when Peter was raised to the rank of department head and was careless enough to leave a portrait of himself floating around." How one man's face came to be plastered all over the Bell Labs campus (including on a watertower) during the early days of Unix development.

Awk is for kids and adults alike, just like A&W's root beer floats.

I misspelled avocado once while grocery shopping. Now I only buy avacardos.

I've seen a few posts like these recently. It's disappointing. In general, I'm a social media lurker, but I'm here because there are some great minds posting interesting things. Without forums like these, there is no good way for me to hear these voices.

New users and organizations can say goodbye to LFS on Hugging Face; Xet is now the default storage for new builders on the Hub 🚀 🚀 🚀 Just sign up for an account, create a new repo, pip install huggingface_hub and you're off! huggingface.co/changelog/xe...

Continuing to move all the LFS bytes into Xet storage on Hugging Face! Currently up to: 🤗 5,500 users and orgs with Xet access 🚀 150,000 Xet-backed models and datasets 🤯 4+ PB managed by Xet How much more to go? If the Hub's top storage users are any indication: many bytes

Qwen is now on Xet on the Hugging Face Hub 🎉 I'm a little star-struck to be honest 🤩 Qwen has 260+ model repos that are regularly among the top trending/downloaded models and we get to serve the bytes from the Hub to your computer.

Before this year, I didn't think about S3.Today, I think about it all the time. Here's a beautiful guest post from Andy Warfield on @werner.social's All Things Distributed about how one engineer (Warfield) came to understand (as much as one can) S3.

Crossed some major milestones in moving @hf.co repos from LFS to Xet: 👷 Over 2,000 builders and nearing 100 organizations with access 🚀 70,000 model and dataset repositories on Xet 🤯 1.4 petabytes managed in Xet storage

This is obvious, but if people broadly rely on a black box to give them answers that help them understand the world, whoever controls the black box controls those answers and therefore the understanding that comes from them. Openness is safety.

The Lovelace 2.0 Test of Artificial Creativity and Intelligence “tell a story in which a boy falls in love with a girl, aliens abduct the boy, and the girl saves the world with the help of a talking cat.” Change that 🐱 to a 🐶 and I'd read that story any day.

The problem with most machine-based random number generators is that they’re not TRULY random, so if you need genuine randomness it is sometimes necessary to link your code to an external random process like a physical noise source or the current rate of US tariffs on a given country.

on the @gradio-hf.bsky.social team, we have been hard at work updating our ImageEditor. You should try it out with this cool model by the brilliant people at Black Forest Labs! huggingface.co/spaces/pngwn...

Xet infra now backs 1000s of repos on @hf.co , which means we get to put on our researcher hats and peer into the bytes 👀 🤓 Xet clients chunk files (~64KB) and skip uploads of duplicate content, but what if those chunks are already in _another_ repo? We skip those too.

Llama 4 dropped this weekend and it was a big one for the Xet team @hf.co - the first major model released entirely on Xet infrastructure end-to-end - more in 🧵

See that purple banner on the Llama 4 models? It's Xet storage, and this is actually huge for anyone building with AI models. Real numbers: ~25% deduplication on Llama 4 models, hitting ~40% for finetunes.

The entire Xet team is so excited to bring Llama 4 to the @hf.co community. Every byte downloaded comes through our infrastructure ❤️ 🤗 ❤️ 🤗 ❤️ 🤗 Read the whole post to see more about these models.

A love letter to the CSV format from @yomguithereal.bsky.social If this doesn't warm your heart, nothing will.

The only thing I do more consistently than write is drink coffee. Yet I will still make these mistakes over and over again. I should add a pre-commit hook for my blog that runs these bash scripts. 🤔