yaroslavvb

8 ms·

31.
by yaroslavvb 7y ago
Divide by 3 to get pre-emptible price
32.
by yaroslavvb 8y ago
It means that there are other factors besides diet. Here's a relevant study https://www.ncbi.nlm.nih.gov/pubmed/707372
33.
by yaroslavvb 8y ago
This kind of scaling was rigorously shown for a related metric called "gradient diversity" in https://arxiv.org/abs/1706.05699
34.
by yaroslavvb 8y ago
I've been using Mathematica since 1995 and Jupyter/colab for 5+ years. Most recently I've been using them both in parallel. While Jupyter is probably the future in terms of mass adoption, there are still some areas where Jupy
35.
by yaroslavvb 9y ago
Underlying implementations often have a way to disable parallelism, ie, OMP_NUM_THREADS=1 or MKL_NUM_THREADS=1
36.
by yaroslavvb 9y ago
You could specify computation in high level API like TensorFlow, and then have the framework pick the best implementations available (ie, CuDNN for GPU, MKL for CPU, something custom for TPU)
37.
by yaroslavvb 10y ago
I found some fonts reuse exact same vector instructions for Cyrillic "А" and English "A" in the same file
38.
by yaroslavvb 11y ago
Vancouver is on the way to becoming the bar from the joke "nobody goes there anymore, it's too crowded"
39.
by yaroslavvb 11y ago
With mean-shift you can get non-convex clusters
40.
by yaroslavvb 12y ago
We found that neural networks can solve CAPTCHAS much better than humans, 99.8% on the "hard" ReCAPTCHA instances: http://arxiv.org/pdf/1312.6082.pdf This is why visual recognition is just one of the signals
41.
by yaroslavvb 12y ago
The most plausible explanation is avalanche: http://en.wikipedia.org/wiki/Dyatlov_Pass_incident#Theories
42.
Russian law requires email providers to keep data in Russia(translate.google.com)
2 points by yaroslavvb 12y ago|0 comments
43.
by yaroslavvb 12y ago
The 99.8% assertion comes from synthetically generated CAPTCHAs. We don't need crowd-sourced ground-truth -- we know what the true answer is because we generated it.
44.
by yaroslavvb 13y ago
"Off-the-shelf" approach didn't work, which is the reason we did this. And the simplifying assumptions are things we could get away with while solving the task at hand. Traditional OCR pipeline would be to use some heuristics
45.
by yaroslavvb 13y ago
It's not really a strawman to say that shuttles take cars off the road because Googlers commuted from the city to Silicon Valley by car before the shuttle program existed. In fact, it started as a 20% project in response to the annoyin
46.
by yaroslavvb 13y ago
By that logic, we should also put some brakes on Bart/Caltrain because it makes it easier to work in Silicon Valley and live in the city.
47.
by yaroslavvb 13y ago
In SF they have permission yet people still protest. Some shuttle stops even have signs which look like they've been installed by the city, ie "No Parking/Shuttle Stop 6am-10am" on Van Ness
48.
by yaroslavvb 13y ago
There's a lot of irrational reactions to shuttles for some reason. I live on top of a "deluxe" cheese shop in Lower Polk, and the owner yesterday told me "Google shuttles, yeah, don't like them, they are causing all
49.
by yaroslavvb 13y ago
Not sure where the controversy is coming from, given that admin clarification from 11-14-2013 said, "You could modify an existing product to integrate with Salesforce and submit that, however you'd be judged on just that componen
50.
by yaroslavvb 14y ago
My dad made an animation with a visual effect in the same spirit: https://plus.google.com/109509141493915423605/posts/jokA4Ccu...
51.
by yaroslavvb 14y ago
It's like "Delta" that makes faucets vs "Delta" that flies planes
52.
by yaroslavvb 14y ago
If you are in Chrome, you can do Incognito Window
53.
by yaroslavvb 14y ago
The tricky thing is that there are hordes of automated captcha-breakers that will recycle captchas until they get something that's easy to OCR
54.
Samsung Galaxy Tab 7.7 Sales Banned in EU by German Court(slate.com)
2 points by yaroslavvb 14y ago|0 comments
55.
by yaroslavvb 14y ago
Oops, looks-like Flume is Google-only name, open-source implementation is called Crunch -- https://issues.apache.org/jira/browse/MAPREDUCE-1849
56.
by yaroslavvb 14y ago
Other technologies to watch: 1. GraphLab2. Unlike Pregel's Bulk Synchronous Parallel Model, GraphLab2 it allows non-synchronous updates, which is more efficient for approximate quantities. For instance, for AltaVista's web graph, most nodes
57.
by yaroslavvb 14y ago
This is not MapReduce though, rather an execution engine specialized for data analysis: http://research.google.com/pubs/pub36632.html
58.
by yaroslavvb 14y ago
They are feeding 10M frames from random YouTube videos, 1 frame per video. Only 3% of 60x60 patches from those frames contained faces
59.
by yaroslavvb 15y ago
Dremel aka BigQuery has a dedicated execution engine, roughly an order of magnitude faster than MapReduce for typical SQL queries
60.
by yaroslavvb 15y ago
Another problem with SIGALARM is that there's only one of them, so setting second alarm will reset the previous counter, and make the system try to run both handlers at the same time on trigger

More ›