Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
-
What the fuck has one of my clients been up to?! Every request he makes he suggests we might want to do it via a function.
"We need to ensure this is password protected. Maybe wrap the calls in a function."
I wander if he goes into the bakery and says, "I need a loaf of bread---use flour."
All I can think of is that someone influential in his life is an FP zealot and he's latched onto this word.
But, I quite like FP. Maybe I'll refactor everything to static classes to meet the requirement.
Hope to Christ he doesn't find out about HTML, etc. "Maybe add a CSS rule."8 -
I had the idea that part of the problem of NN and ML research is we all use the same standard loss and nonlinear functions. In theory most NN architectures are universal aproximators. But theres a big gap between symbolic and numeric computation.
But some of our bigger leaps in improvement weren't just from new architectures, but entire new approaches to how data is transformed, and how we calculate loss, for example KL divergence.
And it occured to me all we really need is training/test/validation data and with the right approach we can let the system discover the architecture (been done before), but also the nonlinear and loss functions itself, and see what pops out the other side as a result.
If a network can instrument its own code as it were, maybe it'd find new and useful nonlinear functions and losses. Networks wouldn't just specificy a conv layer here, or a maxpool there, but derive implementations of these all on their own.
More importantly with a little pruning, we could even use successful examples for bootstrapping smaller more efficient algorithms, all within the graph itself, and use genetic algorithms to mix and match nodes at training time to discover what works or doesn't, or do training, testing, and validation in batches, to anneal a network in the correct direction.
By generating variations of successful nodes and graphs, and using substitution, we can use comparison to minimize error (for some measure of error over accuracy and precision), and select the best graph variations, without strictly having to do much point mutation within any given node, minimizing deleterious effects, sort of like how gene expression leads to unexpected but fitness-improving results for an entire organism, while point-mutations typically cause disease.
It might seem like this wouldn't work out the gate, just on the basis of intuition, but I think the benefit of working through node substitutions or entire subgraph substitution, is that we can check test/validation loss before training is even complete.
If we train a network to specify a known loss, we can even have that evaluate the networks themselves, and run variations on our network loss node to find better losses during training time, and at some point let nodes refer to these same loss calculation graphs, within themselves, switching between them dynamically..via variation and substitution.
I could even invision probabilistic lists of jump addresses, or mappings of value ranges to jump addresses, or having await() style opcodes on some nodes that upon being encountered, queue-up ticks from upstream nodes whose calculations the await()ed node relies on, to do things like emergent convolution.
I've written all the classes and started on the interpreter itself, just a few things that need fleshed out now.
Heres my shitty little partial sketch of the opcodes and ideas.
https://pastebin.com/5yDTaApS
I think I'll teach it to do convolution, color recognition, maybe try mnist, or teach it step by step how to do sequence masking and prediction, dunno yet.6 -
CNC software
Hey guys
what do you guys use to control your DIY CNCs?
I'm using UGS, Universal Gcode Platform, but it has a slight problem... doesn't work with functions.
Meaning, I just had to write a 75 lines program that could be an 8 line program If I could use loops and GOTO...
Also, what CAM software do you use?13 -
Whenever I reach the point where static analysis can't help me any further I always feel a sort of thrill mixed with terror. This is the real deal. Until now the problems were easy to find, the questions had well defined answers to choose from, the rules were universal. In the part of the logic that cannot be checked, the invariants upheld manually, where the best the type system can enforce is for the programmer to clearly state what they're doing, lies the real beast. In proofs commented on functions or invariants as logical expressions over plain English variables written in the doc comments of a struct.
In the blurry and pompous future I imagine for software development, that's where the programmer's time will be spent. Once we all agree on what a string is, what it means to depend on someone else's code, and what parts a UI should be made of, all a developer should have to do is make decisions and derive proofs an automated deduction engine can't do on its own.
Top Tags
Weekly Rant
View