Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Prodigy: A new tool for radically efficient machine teaching

explosion.ai

Post reply on HN
Print view View original

31–40 of 75 posts

Re: Prodigy: A new tool for radically efficient machine teaching

#31

Post by syllogism »

imh wrote: post #20

Since syllogism is participating in this thread, what kind of active learning are you using? I'm always hesitant to use anything except for IWAL since most of the more common ones aren't actually consistent. Even then, then payoff tends to be kinda disappointing. (But I'm definitely not an expert)

Yes, it uses importance weighted active learning. You can set the priorities yourself, but the default built-in sorter is just uses distance from 0.5. There's a random component to help make sure the model doesn't get stuck asking the wrong questions.

Re: Prodigy: A new tool for radically efficient machine teaching

#34

Post by retainingwall »

kafkaesq wrote: post #27
blueyes wrote: post #24

Earlier quoted context omitted.

> spaCY, the leading open-source NLP tool? Sounds like marketing BS. what about OpenNLP and Stanford's for NLP?

spaCY, the leading open-source NLP tool? Agreed, the description is definitely cringe-worthy. As if whoever wrote that wasn't aware that these are language geeks they're marketing to.

Self-respecting language geeks keep up with the times. What's your case for "leading open-source"? Here's a look at Spacy blowing Stanford Core NLP out of the water (via github stars, you can take a look at commits and more from the same tool): https://www.datascience.com/trends?trends=4812,7214,7165&tre...

Re: Prodigy: A new tool for radically efficient machine teaching

#37

Post by aub3bhat »

madisonmay wrote: post #12
En_gr_Student wrote: post #5

So this is just fluff?

Nope, not just fluff. New ML model architectures get too much hype, while it's relatively simple tools like this that actually make the difference in whether or not ML can applied to industry problems. The low hanging fruit in the ML industry are in workflow tools rather than novel model architectures. I have a huge amount of respect for the folks at explosion.ai, largely because their solutions are consistently good…

You might be interested in Deep Video Analytics, its a Visual Data Analytics platform that I am building. [1]

[1] https://github.com/AKSHAYUBHAT/DeepVideoAnalytics

Re: Prodigy: A new tool for radically efficient machine teaching

#38

Post by davedx »

madisonmay wrote: post #12
En_gr_Student wrote: post #5

So this is just fluff?

Nope, not just fluff. New ML model architectures get too much hype, while it's relatively simple tools like this that actually make the difference in whether or not ML can applied to industry problems. The low hanging fruit in the ML industry are in workflow tools rather than novel model architectures. I have a huge amount of respect for the folks at explosion.ai, largely because their solutions are consistently good…

Exactly. I'm working on something related: building a UI on top of declarative ETL pipelines to drive ML models. I think a lot of time (and big data resources) can be saved.

Re: Prodigy: A new tool for radically efficient machine teaching

#40

Post by imh »

syllogism wrote: post #31
imh wrote: post #20

Since syllogism is participating in this thread, what kind of active learning are you using? I'm always hesitant to use anything except for IWAL since most of the more common ones aren't actually consistent. Even then, then payoff tends to be kinda disappointing. (But I'm definitely not an expert)

Yes, it uses importance weighted active learning. You can set the priorities yourself, but the default built-in sorter is just uses distance from 0.5. There's a random component to help make sure the model doesn't get stuck asking the wrong questions.

Thanks for the answer :)

Great work here, btw. It's refreshing to see emphasis on "label some damned data" and work towards making that easy.

Post reply on HN