cdavid

19 ms·

61.
by cdavid 2y ago
For large companies, a big reason is to transform capex into opex, and the predictability. Moreover, large organizations tend to favor predictivity over levels, i.e. are ok to increase average if variance is decreased.
62.
by cdavid 2y ago
It is true both countries were already industrious, but it was far from a given they could go back to their former self. They were both utterly destroyed, and things could have gone really badly, especially in Japan. I can't find the r
63.
by cdavid 2y ago
It is programmable but the HW is sold assembled and ready to use. Wouldn't qualify it as DIY.
64.
by cdavid 2y ago
Not sure whether the word come from there, but sleepers from W. Allen has it in 1973, one of his most fun slapstick movie of that period.
65.
by cdavid 2y ago
In ML, it generally is ordered from most contributor to least contributor, w/ heads of the lab last.
66.
by cdavid 2y ago
It is hard to find modern references on this. When I led a search group, coming from an ML but non search background, I found the following most useful 1. Query understanding series: https://queryunderstanding.com/query-unde
67.
by cdavid 2y ago
If you have access to the query log, aka "who makes which query in what context"), you can use see which queries are "close" to others in context. For example, with session, you can detect manual query rewriting, and use
68.
by cdavid 3y ago
What is your goal ? 1) Know more about how they work on the academic side ? 2) Be able to work in a company that work on LLM ? 3) Be able to work with LLMs 4) agents ? Each of those goal may require different learning "streams" Th
69.
by cdavid 3y ago
Abortion was sensitive in France when it was voted in the early 70ies. Few people in France oppose abortion laws as they are today, but the abortion restrictions are quite different compared to the US. E.g. abortion time limit in France are
70.
by cdavid 3y ago
silos is one of this keyword that is org health dependent. When things go well, people highlight team autonomy. For the exact same setup, when things go bad, people talk about silos.
71.
by cdavid 3y ago
Both Karpathy and fast.ai are good resources. Karpathy is "bottom up" (start from first principles and build on it) while fast.ai is "top down" (you start w/ working examples and gradually "peel off" to un
72.
by cdavid 3y ago
Yes, that's a fundamental reason python packaging is a mess. Python success is largely due to the availability of key mixed language packages. No other mainstream language package manager has to deal with this. For example, cargo for r
73.
by cdavid 3y ago
The question of removing fortran from scipy has happened a few times, and never got anywhere for the reasons you gave. A lot of scipy itself contains fortran code that would take man years to rewrite. Several key parts using fortran have be
74.
by cdavid 3y ago
vespa's blog https://blog.vespa.ai/ will have tons of info about hybrid search
75.
by cdavid 3y ago
If you want, feel free to send me your resume and I can "review" it for you. I've have hired dozens of engineers and EM across ML, BE, etc roles. Info to contact me in my user account (handle + gmail.com)
76.
by cdavid 3y ago
Maybe "risibles amours", but biased as it was my first book from him. As novel, it is hard for me to split between immortality and the joke. The joke touched me more, immortality is more "serious". But hey, I am French,
77.
by cdavid 3y ago
haha, I somehow expected you to participate to this thread. I still disagree with your reading on Kundera, though I found unbearable lightness of being his least interesting work. The joke, "risibles amours" and immortality are gr
78.
by cdavid 3y ago
Middle management (above first line manager) in large organizations is hard to do well. I think it is difficult to understand if you have never done it, and it is even difficult to articulate why. I find it much much harder than working as
79.
by cdavid 3y ago
Not sure I understand your use case, but if you have to use nfs, cachefilesd is very effective for read heavy workload: https://access.redhat.com/documentation/en-us/red_hat_enterp...
80.
by cdavid 3y ago
That somehow surprises me. Can you link such a study ?
81.
by cdavid 4y ago
It releases the GIL only in very specific sections. Most of the Numpy C code runs under the gil. A quick check on master shows only 10-20 calls using NPY_BEGIN_ALLOW_THREADS (which is an alias to Py_BEGIN_ALLOW_THREADS). A lot of the NumPy
82.
by cdavid 4y ago
it is a common misconception that how the retirement is funded makes a huge difference in how much you need to save. In most situations, it does not matter nearly as much as people believe. Dependency ratio will determine in large part how
83.
by cdavid 4y ago
Relying on the cpython gil and releasing the gil are two different thing: 1. releasing the gil means multithreading is opt-in for a given code section in NumPy. Only very specific parts of the code need to be threadsafe. 2. not relyi
84.
by cdavid 4y ago
The obvious one is speed/data size: you can handle much larger data with duckdb compared to pandas. Depending on the data, maybe 10x larger, sometimes even more. From an ergonomics perspective, I find pandas much harder to use casually
85.
by cdavid 4y ago
Depends on the company. Nintendo pays much better than pretty much every game company in Japan. In the last 5-10 years, the gap between Japan and west europe for top tier companies salaries has decreased a lot: you can now get 100k $ or mor
86.
by cdavid 4y ago
He went into politics: https://www.mtdemocrat.com/news/district-1-senate-race-theod...
87.
by cdavid 4y ago
I now old and past 40, but would say it depends on the clubs. There is still quite a lot of new electronic stuff worth listening to. Lots of it is derivative, but that has always been true for most pop music styles. Whatever one may think a
88.
by cdavid 4y ago
It is very context dependent. I am assuming here you mean how to sell your work in your current organization, not in the context of getting a new job. And I am assuming you mean as an individual contributor, and as an engineer (e.g. not as
89.
by cdavid 4y ago
It is not builtin, but ipython will give you that and most features mentioned in the article.
90.
by cdavid 4y ago
That article talks about DL, where blas is much less relevant. The kernels are mostly CUDA (for GPU) and similar stuff for other accelerators.

More ›