julien_c

> Ollama eventually added ollama run hf.co/{repo}:{quant} to pull directly from Hugging Face, which partially addresses the availability problem.

uh actually, _we_ did (generates a Docker-style manifest on the fly)

and a lot of traction on paid (private in particular) storage these days; sneak peek at new landing page: https://huggingface.co/storage

[i work at Hugging Face]

BTW the reporting in this article is sloppy/incorrect

Flamingo-style, see for instance the recently released IDEFICS: https://huggingface.co/blog/idefics

we have a secret plan to change our logo to a Face Hugger some time in the future :)

that would make for a quite different branding.

Might think about it:)

quite cool! haven't tried it yet, but what's the latency on hot-loading a model? (for instance, loading `stabilityai/stable-diffusion-2-1` for the first API call)

what's the inference time on M1?

ArXiv link to the paper: https://arxiv.org/abs/2110.08207

GitHub repo: https://github.com/bigscience-workshop/promptsource

FWIW this is extractive question answering so the output of the model can only be a span of the input paragraph.

SQuAD is the prototypical example of a dataset for this task, see https://rajpurkar.github.io/SQuAD-explorer/

For France, you can check the French Tech Visa program: https://lafrenchtech.com/en/how-france-helps-startups/french...

I've done it for one of my team members, it's pretty easy.

Not really on unsupervised/self-supervised data though, right?

(nor on the same scale of corpora, as far as I can tell)

TL;DR: Hugging Face, the NLP research company known for its transformers library (DISCLAIMER: I work at Hugging Face), has just released a new open-source library for ultra-fast & versatile tokenization for NLP neural net models (i.e. converting strings in model input tensors).

Main features: - Encode 1GB in 20sec - Provide BPE/Byte-Level-BPE/WordPiece/SentencePiece... - Compute exhaustive set of outputs (offset mappings, attention masks, special token masks...) - Written in Rust with bindings for Python and node.js

Github repository and doc: https://github.com/huggingface/tokenizers/tree/master/tokeni...

To install: - Rust: https://crates.io/crates/tokenizers - Python: pip install tokenizers - Node: npm install tokenizers

I had intended to look into Cloudflare before but didn't get a chance. Looks like it's a good time now!

Also https://transformer.huggingface.co

Disclaimer: built it.

Yes it is

Yes, Python's type system is nowhere near Typescript, and are closer to docblock annotations.

We should build one ourselves, actually. That's a terrific idea.

Very nicely said.

The `nn.BatchNorm` CPU inference speedup is a Big Deal™ to us.

Thanks for the contributions everyone, and glad to have bet on PyTorch a few years back.

NYT article from 2016: A Drug to Cure Fear https://www.nytimes.com/2016/01/24/opinion/sunday/a-drug-to-...

Resting rate around 55 for someone who exercises? I don't think I would this bradycardia.

Thanks for the pointer. I can access it here in the US.

Are Techcrunch writers journalists?

No, but it paved the way, in many respects, to those newer approaches – and Ecmascript itself evolving.

Combined with the crazy wind here it also ends up everywhere (like at the top of tall trees)

You can sell reusable plastic bags so if absolutely necessary you can always buy them.

I don't understand why people delete their past Facebook posts. considering it's highly unlikely they are actually deleted in the backend (and not just flagged)

Very timely Show HN post: we just released Mongoku, a neat Web-based interface: https://news.ycombinator.com/item?id=19500859

It scales with your data (at Hugging Face we use it on a 1TB+ cluster) and is blazing fast for all operations, including sort/skip/limit. Built on TypeScript/Node.js/Angular.

[disclaimer: made it]