|
code is hosted at https://github.com/unidef/quantum basically it uses binary trees to high multi dimensional arrays and moves data in that space to simulate a neural network my goal is to write it into the geonode framework
|
# ¿ Jan 16, 2019 23:06 |
|
|
# ¿ Apr 29, 2024 06:42 |
|
next step is header files with inline assembly, but for the next couple of years I’ll need to figure how to move data, so I can do stuff like think(“what’s going on?”, &eax); and neural->y = neural->x; doesn’t work in multiple dimensions, I think
|
# ¿ Jan 16, 2019 23:39 |
|
Dijkstracula posted:
I wrote it that way Incase you have Operation *error_neural_paradiagm() doc *error_neural_doc_system() And you need to change the return types, which could be done in macros. You can also do error_ neural_system()->type according to this Linux kernel book edit: what’s a? that’s 0, 8, 2, 6, hey thanks I think.. hey so uh do you have an referrals on good online math books? or something I can buy on amazon? I failed calculus so bad in college and now I’m a burnout lol unidef freeman fucked around with this message at 10:14 on Jan 17, 2019 |
# ¿ Jan 17, 2019 10:06 |
|
Tankakern posted:but... I know, I learned math from visual studio and Visual Basic when I was a kid and it’s the thrill of electricity and the scene for me, I just have social problems so I can’t really thrive in college I’ve been sending my code to universities and corporations, under the bsd license, but 0 response aside from the sa forums I’ll dig through the git docs, i need to implement a struct with double long integers with arrays for holding data Maybe if Jesus exists he can write me a scheduler for large portions of dma
|
# ¿ Jan 18, 2019 05:26 |
|
Captain Foo posted:Please do not randomly send code to organizations like that I’ve done the latter multiple times, all it did was ruin my life. What’s the worse that can happen? :htp:
|
# ¿ Jan 18, 2019 06:26 |
|
qhat posted:here's a protip: get a name for yourself by actually doing something of value rather than sending random code snippets to people who have no idea who you are Goes the same for you, buddy edit: PEOPLE WILL STEAL YOUR CODE SO BAD IM A loving MORON unidef freeman fucked around with this message at 08:41 on Jan 18, 2019 |
# ¿ Jan 18, 2019 06:41 |
|
DONT THREAD ON ME posted:i'm not sure that i am in on the joke You’re in on it, no ones giving me money and quantum computing is everywhere lol
|
# ¿ Jan 18, 2019 07:28 |
|
DONT THREAD ON ME posted:
Thank you! Be careful with multiple trash dirs and symlinks, I think you can accidentally leak data to other dirs, especially in a source code revisionist system
|
# ¿ Jan 18, 2019 09:18 |
|
DuckConference posted:perfect in every way Thank you! DuckConference posted:perfect in every way Thank you! I need to work in my macro functions, math , and integration into FreeBSD. One goal is to make my code into a pluggable scheduler for hotpluggable memory, maybe for clusters
|
# ¿ Jan 18, 2019 09:21 |
|
gonadic io posted:hello op have you studied the existing work in this space? nottingham uni has people who program with quantum instruction sets on simulated quantum computers: i haven’t done any research in quantum physics.. the quantum name is kind of a monkier for paralellization and neural (believe it or not I coined that apparently after an email to the lkms) space. my goal is to develop a network of “neurons” that you can cluster with binary trees that can quickly read data through memory. now that I think about it I could do something with a cpu caching library, thanks
|
# ¿ Jan 18, 2019 10:29 |
|
gonadic io posted:Let's start with these #includes: include sends the file to the compiler, sys.h is a header file for system macros, structures, variables. the code should be in sys.c sys.h/c is where I’d put platform specific code, like the l1 thing I’m gonna research
|
# ¿ Jan 18, 2019 10:33 |
|
gonadic io posted:i see. i would suggest that you drop the quantum name then because it seems misleading and could leave people thinking that you're a crank. asking about parallel neural networks would get you better responses than with the word "quantum" in there. Well, if you don’t understand how paralellization and neural transparency works in quantum computing, uh,
|
# ¿ Jan 18, 2019 10:49 |
|
fart simpson posted:i just missed my bus stop because i was reading this thread, op Atleast the bus didn’t physically attack you
|
# ¿ Jan 18, 2019 14:03 |
|
Captain Foo posted:hahhaa Well, the compiler gets to read the file, I was oversimplifying. What’s wrong with a sys.h/c file?
|
# ¿ Jan 18, 2019 14:35 |
|
It’s more of a proof of concept of what binary trees can do
|
# ¿ Jan 18, 2019 15:48 |
|
BONGHITZ posted:boi ...bRiGhTaSs? They let you out of the brig? Did you successfully eat the military, you blubbery anime fiend, and hands off my yaoi
|
# ¿ Jan 19, 2019 13:16 |
|
Silver Alicorn posted:op have you tried writing any game mods?? good way to get noticed by Valve I’m not into gaming, but I’d release a GUI that’s Vulkan accelerated Valve is nice.. but I can’t imagine growing older and telling my family I write video games for a living
|
# ¿ Jan 19, 2019 18:13 |
|
Silver Alicorn posted:would you rather tell them you drove for uber for 69 years before dropping dead I’d prefer to grandfather quantum computing, much less work but more stress
|
# ¿ Jan 19, 2019 21:25 |
|
Dijkstracula posted:deeply, cosmically curious as to the meaning of the comment "// prime rib?" While debugging I figured out I could use prime numbers for timing for the threads, forgot how this actually made sense lol
|
# ¿ Jan 21, 2019 11:37 |
|
Don’t be afraid to throw the handle abhja Kalaani around, it’s a fake name
|
# ¿ Jan 21, 2019 11:39 |
|
Pham Nuwen posted:even if it's not intentional, this definitely counts as art: *msg is peliminary debug, the doc type is a buffer of messages
|
# ¿ Jan 23, 2019 00:12 |
|
Made some headway, adding 2d binary tree functions (head/tail) I’m afraid of how much work just throwing a function that allocates a top/bottom at random indices, it reminds me of the hookshoot in wow bfa for rogues
|
# ¿ Jan 23, 2019 21:51 |
|
psiox posted:ship it Give me some time, Xcode ate all my code. Back to emacs, and I’ll put in a push in a week or so, there’s another branch in my GitHub with a few changes
|
# ¿ Jan 24, 2019 00:23 |
|
Captain Foo posted:why is this desirable in an OS scheduler A scheduler organizes application side resources, having a character string attached to each node with updates and debugging resources would be desirable to the Linux world
|
# ¿ Jan 24, 2019 00:53 |
|
Captain Foo posted:are you sure? about any of what you've said here? Lol you sound like Daffy Duck
|
# ¿ Jan 24, 2019 01:04 |
|
Dijkstracula posted:op please address the outstanding questions about // prime rib? itt I think it’s coming back, I was hungry at the time :/
|
# ¿ Jan 24, 2019 03:53 |
|
While we’re prodding, has anyone noticed how monolithic ide’s have become? They used to be intuitive and now there’s a stack of intuition
|
# ¿ Jan 24, 2019 03:54 |
|
Best Bi Geek Squid posted:op you're thinking too small. putting a quantum scheduler in linux would be like strapping a jet engine on a turtle. linux is primitive and newtonian - we need a modern, quantum-first kernel. Well, if you want to know how big I can think I wanted this software to run on spaceships. I’ve said a lot of questionable things, but look at the neural_direction macro, it changes course of the direction the binary trees are pointing. I think it was my cousins iranian friend said there’s potentional for nuclear code in there!
|
# ¿ Jan 24, 2019 19:07 |
|
Dijkstracula posted:op please address the outstanding questions about // prime rib? itt it’s uhm coming back, I was thinking of dinner. goon power
|
# ¿ Jan 24, 2019 20:24 |
|
JawnV6 posted:that doesnt make any sense? Think a free Unix to be run on quantum mechanics, think of this, there’s a grid of pointers, a function reads the quantum_direction macro and the matrice of pointers moves in that direction. This could be useful for hyperspace
|
# ¿ Jan 25, 2019 00:03 |
|
Tomorrow I’m going to try to write some code in Todo: Find out whatever *malloc() returns and create a proper function (or sadly a macro) that allocates a neuron Figure out this multidimensional binary tree (where and how do I implement multi heads and multi tails? Please god no macros, I’m thinking of a multidimensional array) Build tools for database management (including random characters generator to fill up databases in ram)
|
# ¿ Jan 25, 2019 05:31 |
|
Captain Foo posted:what's wrong with /dev/urandom The tool is mainly for benchmarking, and this isnt a sql database, it’s something concoceted by me
|
# ¿ Jan 25, 2019 05:39 |
|
Dijkstracula posted:op why do you care so much about macros I think it goes back to when i was a kid, and my dad took me to work, and a macro killed my dad
|
# ¿ Jan 25, 2019 06:57 |
|
qhat posted:if only sons were like fathers ah the #define keyword, come to finish the job, eh?
|
# ¿ Jan 25, 2019 07:42 |
|
didn’t see the stealth edit
|
# ¿ Jan 25, 2019 08:13 |
|
elite_garbage_man posted:you should post this update to the linux mailing list thread. When I finish my small library I’ll give it a go at implementing it inside the Linux kernel, then send a patch
|
# ¿ Jan 25, 2019 16:36 |
|
gonadic io posted:don't worry op, in your case (and in fact all cases) malloc won't return anything important you can do void to void but I settled with a macro
|
# ¿ Jan 27, 2019 00:21 |
|
new push on https://github.com/unidef/quantum I kind of gave up, will work on this project on and off could someone fix my code? is the data structure supposed to go inside the node struct or visca versa?
|
# ¿ Jan 30, 2019 03:40 |
|
qhat posted:Op your binary graph has too many matrixs, it can't fit into the cache lines on your cpu without multifactored prefetching. Awha? You can make binary graphs out of c code? Try adjusting the dimension macro to about 5 for modern hardware
|
# ¿ Jan 30, 2019 04:17 |
|
|
# ¿ Apr 29, 2024 06:42 |
|
shopping list gin (mornings) whiskey (evenings) rum (nights-mornings) just my regime for dissecting the Linux kernel
|
# ¿ Jan 30, 2019 13:07 |