gtani

17 ms·

121.
by gtani 10y ago
There was a thread recently, maybe on reddit, about a facebook group or somewhere people give away ec2 or azure credits to people who want to do HPC or deep learning or whatever. But i can't google it.
122.
by gtani 10y ago
Tim Miller, now in Carlsbad. There's lots of good teachers in LA area.
123.
by gtani 10y ago
There was an ashtanga teacher in SoCal who did pranayama practice with a few students prior to Mysore practice. It was really demanding, somebody counted 55 inhale-exhale cycles in an hour, something like that
124.
by gtani 10y ago
Yes, the quad cores sold prior to Oct 2014 can be found in craigslist/ebay etc for US$700-1100 tho i think they're limited to 16Gb RAM http://barefeats.com/macmin14.html
125.
by gtani 10y ago
His previous slide decks, which seem to be structured a little differently (I've only looked for ~5 minutes) are available, so I expect this one will be too http://www.cs.cmu.edu/~rsalakhu/ http://www.c
126.
by gtani 10y ago
recent discussion of running tensorflow/grad. descent on DSP chips: https://news.ycombinator.com/item?id=13379847 This is a good writeup on mobile/low power devices, lower precision arithmetic, compression, sparsi
127.
by gtani 10y ago
did you try http://nlp.stanford.edu/projects/coref.shtml
128.
by gtani 10y ago
Agreed, as well as the evolving nontrivial hunt for cost reductions e.g. "Just use spot instances in US west 1 on weekends", stuff like that. A pascal GPU and older CPU that takes at least 32GB RAM is a decent combination. I'
129.
by gtani 10y ago
Besides the DL book, there's other excellent texts that are freely available/open content on the web: - Elements statistical Learning, Hastie et al - Shalev-Shwartz and Ben-David: http://www.cs.huji.ac.il/~shais&#x
130.
by gtani 10y ago
One thing about some of those elite institutions, Columbia, Penn, Chicago, Yale, teh area south of UC-Berkeley, are that they find themselves in the middle of not well doing urban centers. You could call them ghettos. You have a lot of pr
131.
by gtani 10y ago
There's other classics by Martin Russ, Miller Puckette, Curtis Roads, Allen Strange, Perry Cook.
132.
by gtani 10y ago
We've had a bunch of really good threads, when i have time i'll stick them in pinboard.in: Faust: https://news.ycombinator.com/item?id=13012880 audioKit: https://news.ycombinator.com/item?id=990376
133.
by gtani 10y ago
I'll try to read this research note. Maybe I'm biased. I used to work at Moody's and then 2 or so jobs later structured some pretty substantial bond issuances that they rated (and my manager at that gig came from S&P, a
134.
by gtani 10y ago
If you're getting a single video card, or one for calculation and one just to do video, make sure motherboard has 16 PCI-e rel3 lanes into the primary card, people are using z170 and z97. For 4 video cards, you want x99 mobo, newegg d
135.
by gtani 10y ago
Here's the current water content of snowpack, far below normal: http://cdec.water.ca.gov/cdecapp/snowapp/sweq.action
136.
by gtani 10y ago
Knobby synth (ms2k, sh201, jp8k or 8080, Supernova, ESQ1, DW6k/8k, AX60, KS4, AN1X etc etc) + knobby/easily mapped controller (i use Axioms and Novation SL Mark 2 even tho the pots/faders/encoders have proved not especia
137.
by gtani 10y ago
Wow, that's a lot of good work. I asked recently about go wrappers for the Cuda Runtime or Driver API's and didn't get much response, I don't see a lot happening with openCL either, is there anything like that in the pro
138.
by gtani 10y ago
I'll try this later, seems to be targetting pre-Pascal cards (need more info on whether this will work with Cuda 8). "Cuda for Engineers" (excellent book by Storti/Yurtoglu) uses VS 2013 and the VC++ that comes with it,
139.
by gtani 10y ago
I think this is it: http://mra.org/wp-content/uploads/2016/05/Beverly_IceScrews_... It's been a while since i've thought about ice climbing, but i think most/many ice climbers are using t
140.
by gtani 10y ago
Did you reach out to Nvidia? It seems like they should be willing to help with putting out tested/documented bindings for more languagues. Also, where in the repo is your wrapper? (I don't have enough knowledge to debug s.t. like
141.
by gtani 10y ago
Fixed up GPU work I haven't thought about using golang to feed data to a GTX 1080 or similar, but before i do that, i'd like some info on the above, and a quick google shows that there may not be a complete wrapper for t
142.
by gtani 10y ago
Interested to see where this goes. The 5 blogs/books cited at the end look especially interesting. I never thought about music notation until i learned guitar. Previous to that i was "gifted" years of lessons in piano, wood
143.
by gtani 10y ago
There's some free VST's that are supposed to be good, and reaper, which isn't really free, but the trial period is at your discretion. Before you start with a DAW, read lots of reviews, watch youtube walkthrus and ask around
144.
by gtani 10y ago
For those interested in online/soft synths, sequencers and samplers (pretty deep rabbit hole)look up past discussions of Max for Live, ChucK, SuperCollider, Impromptu, CSound. Max Live is part of Ableton Suite license for around $700,
145.
by gtani 10y ago
I learned C and IBM APL2 around the same time for financial apps. For APL, i read the Polivka and Pakin book about 30 times, not this edition, but the red paperback I used is apparently impossible to buy. Excellent book, it got me to wher
146.
by gtani 10y ago
There's a few projects that use ES/lucene as a backend/datastore once the feature engineering is done, but I don't see models operating on the native indexes directly, maybe the format is too different from one-hot (afte
147.
by gtani 10y ago
I think the comment to use one of successor languages is most pragmatic. I learned with stickers on keyboard, you don't want to do that, and a dedicated keyboard sounds kind of silly (this from former pro IBM APL2 programmer, it was m
148.
by gtani 10y ago
Well, there's not much price pressure from azure, you could try the others in nvidia's list: https://www.reddit.com/r/MachineLearning/comments/46ict5/bes... Also watch spot price instances, go
149.
by gtani 10y ago
People commonly spend US$80+/month on training deep learning models on EC2, when you can now buy a brand new 980Ti for $440 from newegg, and the very capable 970 at $280. If you don't have a motherboard/PS that supports 2 vi
150.
by gtani 10y ago
this sub's wiki is pretty comprehensive: https://www.reddit.com/r/MachineLearning/wiki/index If you want to be taken seriously, you need to learn calculus, linear algebra, convex optimization, probabilit

More ›