etrain

8 ms·

1.
by etrain 4y ago
There was a company called Petridish that did exactly this. They pivoted after six months and became Blue Apron: https://vator.tv/news/2017-09-05-when-blue-apron-was-young-t...
2.
by etrain 5y ago
Users expose their model to our Trial API ( https://docs.determined.ai/latest/topic-guides/model-definit... ), the base class then implements a training loop (which can be enhanced with user-supplied callbacks, metr
3.
by etrain 5y ago
Check out Determined https://github.com/determined-ai/determined to help manage this kind of work at scale: Determined leverages Horovod under the hood, automatically manages cloud resources and can get you up on spot
4.
by etrain 6y ago
Tom7’s sigbovik submissions are true works of art. https://www.cs.cmu.edu/~tom7/mario/mario.pdf
5.
AI infrastructure for everyone, now Open Source(determined.ai)
12 points by etrain 6y ago|0 comments
6.
by etrain 6y ago
The exposition is tight.
7.
by etrain 7y ago
Relational database queries are supposed to be _declarative_. As a user, you're not supposed to think about the mechanics of execution because the database system is supposed to be able to decide how to execute your query using whateve
8.
by etrain 7y ago
Assume 100 pages on each onion address (it’s probably power-law but let’s just assume that’s the mean). Latency with Tor is super high. Assume average of 5s to load a single page. This is generous because tail latency will probably domina
9.
by etrain 8y ago
Yes, this is correct: http://composition.al/lindsey_kuper_cv.pdf
10.
by etrain 8y ago
It’s also hard to separate the design of the neural architecture from the definition of the feature extractor.
11.
Why Reproducibility Matters in Deep Learning(determined.ai)
3 points by etrain 8y ago|1 comments
12.
by etrain 9y ago
I had one of these phones too - it was a bad batch of serial numbers according to Apple. If this happens to you: https://www.apple.com/support/iphone6s-unexpectedshutdown/
13.
by etrain 9y ago
Crunchbase has a good deal of data here - it is incomplete particularly for smaller companies, but is pretty good for larger investment rounds.
14.
by etrain 10y ago
A word of caution - these talks are highly technical and theoretical and I don't recommend them for anyone looking to understand practical applications of ML. If, however, you're interested in intro to cutting edge work at the fro
15.
by etrain 10y ago
We've used this dataset to build a product review classification pipeline as an example application that can be developed using our project, KeystoneML (which runs on spark) - code is here: https://github.com/amplab&#x2
16.
iPhone 6s Program for Unexpected Shutdown Issues(apple.com)
2 points by etrain 10y ago|0 comments
17.
by etrain 10y ago
You might find the OSDI slides useful: https://people.eecs.berkeley.edu/~apanda/assets/slides/osdi-...
18.
by etrain 10y ago
I would also be interested in seeing this.
19.
by etrain 10y ago
Microsoft is 23 years older than Google and it usually takes time for a corporation to create an R&D lab. Given Google's rich academic roots this happened faster there than usual. Google is doing a good job given that it's a r
20.
by etrain 10y ago
This is pretty awesome. One key bit of information that the compiler has is that the coefficients are a constant array of length 12, which makes the loop unrolling possible and also means that the register magic is in play - it's serio
21.
by etrain 10y ago
GPUs in the cloud aren't targeted at gamers. They're targeted at people doing things like running render farms and training deep learning models.
22.
by etrain 10y ago
Keep in mind that with EC2 you're billed hourly so the fastest a 70 cpu-hour job could finish on m3.medium for $0.70 is 1 hour, and that's ignoring setup time, etc. Meanwhile, on Lambda, you can actually run 1600 60s jobs (or 27
23.
by etrain 10y ago
One of the nice things about lambda is the billing is super granular - you get billed at 100ms intervals. Assuming 70 hours at $0.000000834/100ms [1] The whole job costs $2.10. [1] - https://aws.amazon.com/lambda&#x2F
24.
by etrain 10y ago
This varies highly depending on the organization. I've worked at places that do it both ways. Particularly in small places when resources are tight, teams are distributed, or there is a non-code safety measure in place (e.g. Business d
25.
by etrain 10y ago
If I were trying to eat that diet and not break the bank/use all my time I'd invest in 4 things: 1) A ricecooker. 2) A food processor. 3) A sous-vide/vaccum sealer. 4) A Costco membership. The ricecooker makes rice and keeps
26.
by etrain 10y ago
If you're willing to try something more bleeding edge: http://btrdb.io/
27.
by etrain 10y ago
type systems are great.
28.
by etrain 10y ago
Going rate for humans breaking captchas is like $1/1000 captchas solved. A fully automated service could make some money, sure, but not exactly a killing .
29.
Music classification in a Tweet(eecs.berkeley.edu)
5 points by etrain 10y ago|0 comments
30.
Saddles again(benjamin-recht.github.io)
3 points by etrain 10y ago|1 comments

More ›