5 ms·

Abandoned, but large parts of this seem to be the work of Yann LeCun (https://www.facebook.com/yann.lecun/posts/10151728212367143 https://www.facebook.com/yann.

by nnq 11y ago

Abandoned, but large parts of this seem to be the work of Yann LeCun (https://www.facebook.com/yann.lecun/posts/10151728212367143 https://www.facebook.com/yann.lecun/posts/10151728212367143), Facebook's director of AI research, right? ...curious if scraped parts of this project are actually living a second life deep inside Facebook's labs :)

albertzeyer 11y ago
And Bengio and some other well-known people from the Deep Learning community. Interesting.
brudgers 11y ago
I sometimes wonder if there is too much emphasis on the idea of that some piece of software is abandoned. By which I mean that new and actively developed software isn't bug free and if a tool does the job it does the job. There's no guarantee that an active project in version 0.9 is going to fix any particular bug soon short of me [in theory] fixing it myself. It's not as if much has changed in 2015 that would introduce show stopping bugs that weren't around in 2009. The process is the same, evaluate the tool against the problem domain and compile from source if it seems like a reasonable fit. Active development doesn't change the fact that there is no silver bullet.
- masukomi 11y ago
  You're right in general, but it seems that no language that is being actively used by a non-trivially sized community ever sits still. The more we use our languages (computer or human) the more they evolve. They are a form of communication. We are always evolving our communications methods and wanting more ways of expressing ourselves.
  - brudgers 11y ago
    1. ANSI Common Lisp is nearly 25 years old. 2. Picking up another language is the quantum leap in the possibilities of expression. Point releases aren't. To the point, if Lush expresses what a person wants to express, then it does so. It's not like people suddenly turn around and rewrite working legacy while loops as list comprehensions just for fun.
ihnorton 11y ago
His lab now uses (and contributes to) Torch7: http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_yann_lecun/chisdw1 http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_...
smhx 11y ago
I used Lush for my homework as part of the Machine Learning course at NYU (taught by LeCun). It was pretty mature, mostly bug-free for general use. However, at that time, I never used a lisp-y language before that, and found it quite difficult to do my homeworks. It had an interesting mode where if you annotate your lush source-code with type-hints, all your Lush transpiles in to C, which you can compile with your favorite C compiler. It was eventually abandoned in usage because Torch (Lua based scientific-computing platform) was adopted by the lab. AFAIK, Lush is not used inside Facebook labs.

by nnq 11y ago

albertzeyer 11y ago

And Bengio and some other well-known people from the Deep Learning community. Interesting.

brudgers 11y ago

I sometimes wonder if there is too much emphasis on the idea of that some piece of software is abandoned. By which I mean that new and actively developed software isn't bug free and if a tool does the job it does the job. There's no guarantee that an active project in version 0.9 is going to fix any particular bug soon short of me [in theory] fixing it myself. It's not as if much has changed in 2015 that would introduce show stopping bugs that weren't around in 2009. The process is the same, evaluate the tool against the problem domain and compile from source if it seems like a reasonable fit. Active development doesn't change the fact that there is no silver bullet.

masukomi 11y ago
You're right in general, but it seems that no language that is being actively used by a non-trivially sized community ever sits still. The more we use our languages (computer or human) the more they evolve. They are a form of communication. We are always evolving our communications methods and wanting more ways of expressing ourselves.
- brudgers 11y ago
  1. ANSI Common Lisp is nearly 25 years old. 2. Picking up another language is the quantum leap in the possibilities of expression. Point releases aren't. To the point, if Lush expresses what a person wants to express, then it does so. It's not like people suddenly turn around and rewrite working legacy while loops as list comprehensions just for fun.

ihnorton 11y ago

His lab now uses (and contributes to) Torch7: http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_yann_lecun/chisdw1 http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_...

smhx 11y ago

I used Lush for my homework as part of the Machine Learning course at NYU (taught by LeCun). It was pretty mature, mostly bug-free for general use. However, at that time, I never used a lisp-y language before that, and found it quite difficult to do my homeworks. It had an interesting mode where if you annotate your lush source-code with type-hints, all your Lush transpiles in to C, which you can compile with your favorite C compiler. It was eventually abandoned in usage because Torch (Lua based scientific-computing platform) was adopted by the lab. AFAIK, Lush is not used inside Facebook labs.