ulrikhansen54

10 ms·

31.
Show HN: Open-source toolkit for ML model evaluation and active learning(github.com)
11 points by ulrikhansen54 3y ago|0 comments
32.
by ulrikhansen54 3y ago
I'd download and run a few open-source LLMs that can be run without internet access and a laptop - can be an infinite source of math problems, quizzes, programming challenges, etc., that should keep you occupied for a very long time.
33.
Fine-tune the Segment Anything foundation model: A step-by-step guide(encord.com)
7 points by ulrikhansen54 4y ago|0 comments
34.
by ulrikhansen54 4y ago
I bloody hope so...
35.
Playboy: The Bill Gates Interview (1994)(tech-insider.org)
3 points by ulrikhansen54 4y ago|0 comments
36.
by ulrikhansen54 4y ago
'Unsupervised reinforcement learning' is how these large models and systems ultimately will end up becoming sentient. We recently tried a similar approach on a toy problem in the computer vision sphere ( https://encord.c
37.
by ulrikhansen54 4y ago
Segment Anything is orders of magnitude more powerful than anything we've ever seen before within the real of image segmentation and the first 'foundation model' in that regard, so this is a real step-change in background rem
38.
by ulrikhansen54 4y ago
They pull everything from various e-print archives (arXiv, ResearchGate, etc.) AFAIK Edit: Everything tagged with AI/ML
39.
by ulrikhansen54 4y ago
https://paperswithcode.com/ is arguably the best source and overview of all the research. Its also (somewhat) unbiased (owned by Meta), not being an SEO-optimised company blog.
40.
by ulrikhansen54 4y ago
Ty for flagging -- computer vision just got a whole lot easier :-)
41.
Meta AI's Segment Anything Model (SAM) explained(encord.com)
17 points by ulrikhansen54 4y ago|2 comments
42.
The Real Competition Is the Water(rkg.blog)
2 points by ulrikhansen54 4y ago|0 comments
43.
by ulrikhansen54 4y ago
It's not that there's only 10% left to improve. It's that the data needed, compute requirements, and model size are as intensive, getting from 0 to 80 as they are getting from 80 to ~85 or ~90. See https://paperswi
44.
by ulrikhansen54 4y ago
Ha, touché...
45.
by ulrikhansen54 4y ago
There's about ~10% point improvement left (i.e, from 80% to 90%) before it starts to stagnate. We've seen the same with predictive models benchmarked on ImageNet et. al.
46.
by ulrikhansen54 4y ago
New ones will emerge (eventually)
47.
by ulrikhansen54 4y ago
These guys are smart
48.
Micro-models: purposefully overfit models that are good at one specific thing(eric-landau.medium.com)
72 points by ulrikhansen54 5y ago|30 comments
49.
Cord raises $4.5M to automate computer vision annotation processes(venturebeat.com)
2 points by ulrikhansen54 5y ago|0 comments
50.
by ulrikhansen54 6y ago
Can you shoot us an email at ulrik.hansen@cord.tech -- we can continue the conversation there.
51.
by ulrikhansen54 6y ago
Would love that. Can you shoot me an email: ulrik.hansen@cord.tech
52.
by ulrikhansen54 6y ago
To my understanding Aquarium is focused on curation of datasets rather than labeling (e.g. finding edge cases in data, etc.) and have built an awesome product around that. Aquarium works well in domains that are further along on the ML adop
53.
by ulrikhansen54 6y ago
Looking forward to it!
54.
by ulrikhansen54 6y ago
Thanks for sharing those evilolive! I haven't come across the Toronto Univ annotation suite before --- polygon auto-segmentation is an interesting problem. We have embedded a simple auto-segmentation algorithm into our GUI that run wit
55.
by ulrikhansen54 6y ago
Thanks! We've really enjoyed engaging with Lenus.
56.
by ulrikhansen54 6y ago
Yes! We do on premise deployments for enterprise. But even with our cloud solution, our philosophy and terms of services is that all your data and any models you build through the platform are 100% yours. There are some companies that are s
57.
by ulrikhansen54 6y ago
Thanks! We don't have any immediate plans to expand into other data types. We've decided to remain focused on just one 'data vertical' (anything computer vision related) for the foreseeable future, as we want to build ou
58.
Launch HN: Cord (YC W21) – training data toolbox for computer vision
64 points by ulrikhansen54 6y ago|25 comments
59.
Why You Should Ditch Your In-House Training Data Tools(ulrik-hansen.medium.com)
2 points by ulrikhansen54 6y ago|0 comments
60.
by ulrikhansen54 6y ago
That's the idea

More ›