PHAnna_sarap69
Honorary Poster
Mga kuyas gusto ko gumawa ng AI. to stop humans abusing planet earth po. Ano po una gawin?
-
Welcome to PHCorner Forums! Sign up for free to unlock full access and exclusive privileges, including:
- Asking questions about computer and mobile issues
- Discovering internet tips for your mobile network and computer
🔒 Closed Computerchess free softwares for learning, training, testing....and everything you need to know PART1
- Thread starter alist1986
- Start date
- Status
. Mag-exit na sana ako he he, pero muntik ko nang ibuga yung kape sa bibig ko ng mabasa ko ito. Baka sa "Greenpeace" site meron nyan. Bakit mo naisip yan, sa dinami-dami ng pwedeng isipin?Sumali ka dito: You do not have permission to view the full content of this post. Log in or register now.
Basahin mo ito at related sa tanong mo: You do not have permission to view the full content of this post. Log in or register now.
Ang bilin ko lang, huwag mong babasahin yung mga AI comments ni Stephen Hawking at panaginip lang iyon he he.
tolyakarpov
Forum Veteran
Kung ganun mag-clone ka na lang at magreverse engeenering ng chess engine..tapos pa test mo sa amin...para maeron tayong phcorner chess engine
tolyakarpov
Forum Veteran
Marming volunteer na tester ang leela chess zero eh kaya sa maikling panahon lumakas agad,dapat siguro magtest at magtrain ka rin ng leela zero,
Hindi na kailangan dahil matakaw sa pc time yang training. Online service yang ginagawang training para continuous yung pag-gather ng data para sa neural network. Kaya ako tumigil noon dahil 2 pc ang gamit ko para lang sa CE testing, equivalent to +3K electrical bills a month na free service lang for CCRL testing group. Buti noon, kumikita ako ng malaki kapag may contract work ako abroad. Ngayon, wala na.
Pag-aaralan ko muna kung may paraan na mag-edit or merge ng neural network para madali. Di ko pa alam kung meron din paraan para ma-filter yung files na yan.
Para yang chatbot din, na tinuturuan mong mag-converse near human level by adding experience files. Pag may oras, pwede kong gawin yan.
Madaling mag-clone. Gusto ko man, di ko alam kung may oras ako para dyan. Malakas kasing kumunsumo ng kuryente ang chess engines sa pc - parating 100% cpu usage kung lahat ng cores ay gagamitin. Iniingatan ko na yung last laptop ko't baka bumigay he he. Pero sa minor coding/compiling ay pwede naman. Wala na yung dati kong tools for reverse engineering at yung dati kong kakilala para basahin yung raw code to convert to standard coding language. Baka dedbol na he he at matanda pa yon sa'kin. Sa opening book na lang ako mas interisado kaysa sa chess engine modding. Yan yung project ko ngayon para ilaban sa Sedatchess opening book tournament kung ok yung results. I-release ko as new RodentChess book kung gusto ni Pavel Koziol. Matagal kasing mag-analyze ng games to build as opening books. Kapag nainip ako, pinapatay ko muna dahil di lang ako yung gumagamit ng pc sa bahay.Kung ganun mag-clone ka na lang at magreverse engeenering ng chess engine..tapos pa test mo sa amin...para maeron tayong phcorner chess engine
Ang best method for non-programmers is to learn how to tune the Chess Engine UCI parameters to it's optimum potential. Yan yung importanteng makuha ng gumagamit, dahil yung default ay hindi 100% perfect. Tapos, hahanapin ng user yung appropriate opening lines suited for the engines technical theory for perfect play. Para ito sa mga lumalaban ng CE vs. CE only online. Iba yung to play against humans. Yan kasi yung ginagawa ko noon.
Maganda sana kung meron kayong pc rin at yon ang platform na gamit ko. Merong mga source codes for chess engine analysis ng games na nabuo ko na sa Windows na pwede ninyong i-test din. At maraming rin less known softwares para dyan sa github.com na mas madaling gamitin. Yan yung iniipon ko at binubuo ng dahan-dahan para makumpleto ko yung gusto kong chess engine utilities for Windows. Di ako palagamit ng commercial softwares at sa open source ako laging nakatuon.
Sa python ay meron din direct sa console na applications. Di ko alam kung pwede yon straight sa android. Durable naman ang python language.
RomanRapido
Grasshopper
Pa advice sana kung ano yung the best na endgame trainer. thanks!
Spoiler contents are visible only to Established Members.
Anna_sarap69
Honorary Poster
Kuya sa android parang alang chance na lumakas ang leela chess zero?
Anna_sarap69
Honorary Poster
Wowowin po tayo dito?
tolyakarpov
Forum Veteran
Depende sa weight yan..kuha ka ng mastaas na weight.kung tao lang naman lalaban naku malakas na rin.
tolyakarpov
Forum Veteran
Lahat naman ng chess engine na may elo na 2800 pataas ang matalas na sa endgame.ang gamit ko ay stockfish 11,shash chesss at komodo 10.
tolyakarpov
Forum Veteran
Hindi pala ganun kadaling gawin.tapos magastos pa pala.Pangmayaman pala na hobby ito.
Isipin mo lang na walang patid na nakabukas yung pc mo na nasa 100% cpu usage, malaking gastusin din yon di ba? Yung testing kasi ay 24/7 to update the ELO ranking sa mainsite. Libu-libong games kasi yung yung CE-testing. Ang advantage nga lang ay nabibigyan ka ng early releases ng commercial engines for free (kung galante yung author) at mga beta releases ng open source engines na di pa released.
Buti ngayon may cloud testing, pero pera din yon dahil magbabayad ka ng online account to use their service.
Kaya mas mainam na lang yung pangsarili mo, na depende sa necessity mo sa paggamit.
Pangmayaman yan kung sakaling ginagastusan mo rin ng state-of-the-art hardware like cluster cpus, latest multi-CPU or multi-GPU hardware kada may bagong release sa market. Yung Go, Othello, at 12x12 checkers at iba pang board game AI engines ay ganyan din.
Buti ngayon may cloud testing, pero pera din yon dahil magbabayad ka ng online account to use their service.
Kaya mas mainam na lang yung pangsarili mo, na depende sa necessity mo sa paggamit.
Pangmayaman yan kung sakaling ginagastusan mo rin ng state-of-the-art hardware like cluster cpus, latest multi-CPU or multi-GPU hardware kada may bagong release sa market. Yung Go, Othello, at 12x12 checkers at iba pang board game AI engines ay ganyan din.
Parang sinagot ko na ito noon ah. Ba't nawala he he. Eto na lang yung tandaan mo bosing. Kung sa android mo gagamitin ang LC0 at engine opponents, kahit pa 4-cpu yan at pinakamalakas na neural network ang gamit mo na kaya ng cpu ng android mo (dahil hindi lahat ay pwede mong gamitin or else ay mag-hang yan), hindi niya kayang talunin yung "single core" SF or SF clone. Known fact na yan. Wala rin namang malakas na GPU yung typical android device, kaya expect na hindi yan yung expected mo na LC0 strength kung sa CPU mo lang ng android gagamitin. Test mo rin kung saan malakas yung LC0 - sa Eigen, OpenBlas or OpenCL releases.
You do not have permission to view the full content of this post. Log in or register now.
Yung nababasa mong high strength ng LC0 ay kung gagamitin siya sa isang malakas na GPU, TPU. Dyan lang may advantage ang LC0. Just accept what your device can provide for LC0.
Tama rin yung sinabi ni tolyakarpov . Try to use better neural network para lumakas siya. Dyan naman talaga naka-rely yung lakas ni LC0.
Check mo yung recommendations dito, para tama yung classification ng weights na gagamitin mo for the capacity of your android (CPU or GPU) device - 10b (10 blocks) pababa for android, pero nasa iyo na kung subukan mo sa higher blocks/filters. Ex. 9 blocks x 112 filters then look for best 10 blocks x 128 filters. Hindi kasi lahat yan ay stable gamitin dahil baka mag-hang lang si LC0, sobrang tagal mag-isip or mas lalong humina he he.
I recommend this for most android devices:
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
Yung nababasa mong high strength ng LC0 ay kung gagamitin siya sa isang malakas na GPU, TPU. Dyan lang may advantage ang LC0. Just accept what your device can provide for LC0.
Tama rin yung sinabi ni tolyakarpov . Try to use better neural network para lumakas siya. Dyan naman talaga naka-rely yung lakas ni LC0.
Check mo yung recommendations dito, para tama yung classification ng weights na gagamitin mo for the capacity of your android (CPU or GPU) device - 10b (10 blocks) pababa for android, pero nasa iyo na kung subukan mo sa higher blocks/filters. Ex. 9 blocks x 112 filters then look for best 10 blocks x 128 filters. Hindi kasi lahat yan ay stable gamitin dahil baka mag-hang lang si LC0, sobrang tagal mag-isip or mas lalong humina he he.
I recommend this for most android devices:
Spoiler contents are visible only to Established Members.
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
Naku, kahit huli na ay balikan ko ulit yung tanong mo bosing dahil nabago ang ihip ng hangin he he.
Sorry po. Meron palang released na "Chess Alpha Go" bago nagkaroon ng Leela Chess Zero. Nakita ko lang dahil naghahanap ako ng projects sa You do not have permission to view the full content of this post. Log in or register now..
Ito yon, yung parehas na ginamit ng Deepmind noon versus Stockfish though TPUs yung hardware nila:
You do not have permission to view the full content of this post. Log in or register now.
Kung may malakas kang GPU sa pc mo like "GEForce GTX 1050 man lang" ay uubra din. Ang alam ko kasi, below this model ay maraming problema kahit sa LC0.
Install ka ng Tensorflow 2, Python 3.63 or newer version, Keras 2.0.8 - nandoon naman lahat yan sa python requirements, then follow the instructions.
But best is just use LC0 dahil halos walang pinakaiba yan sa Chess Alpha Go - marami pang platforms para magamit siya bukod sa auxiliary files (weights, training data, etc.).
Nagbaka-sakali lang ako to those who maybe interested to test and start from scratch.
OK. Unahin ko na lang muna yung sa basic training - beginners and up. Be sure alam mong mag-fix thru Windows compatibility since luma yung mga ito kung sakali.
You do not have permission to view the full content of this post. Log in or register now.
You do not have permission to view the full content of this post. Log in or register now.
Baka kasi Win10 yang OS mo.
BASIC FREE CHESS TRAINING SOFTWARES:
Punta ka dito: Free lahat yung nandyan at buhay pa "lahat" yung softwares. Checked ko na yan ngayon.
You do not have permission to view the full content of this post. Log in or register now.
BASIC FREE CHESS GUI AND DATABASE SOFTWARES:
1. ARENA CHESS GUI - You do not have permission to view the full content of this post. Log in or register now.
2. SCID Chess Database Software - You do not have permission to view the full content of this post. Log in or register now.
3. Scid VS. PC Chess Database Software - You do not have permission to view the full content of this post. Log in or register now.
4. CHESSX Chess Database and PGN viewer - You do not have permission to view the full content of this post. Log in or register now.
5. PGN Chessbook Chess Database and PGN viewer - You do not have permission to view the full content of this post. Log in or register now.
Best free sources of annotated pgn for basic training/study:
1. You do not have permission to view the full content of this post. Log in or register now.
2. You do not have permission to view the full content of this post. Log in or register now.
3. You do not have permission to view the full content of this post. Log in or register now.
4. You do not have permission to view the full content of this post. Log in or register now. - hanapin mo na lang dyan (best site ito)
For normal pgn and opening line review (players/eco lines), just start here;
You do not have permission to view the full content of this post. Log in or register now.
Check mo rin yung iba kong posts dito dahil lahat naman yon for PC, including Chessbase/Fritz- the commercial ones.
Anything related to computer chess, dito ka mag-umpisa to be well-informed:
You do not have permission to view the full content of this post. Log in or register now.
Then visit this site:
You do not have permission to view the full content of this post. Log in or register now.
I-link ka nyan sa mga kailangan mong malaman - chess engine links/downloads, general interests, tournaments, utilities, etc...
Kapag nakapa mo na, say yung Arena, then we go to opening books/epd/test suites. etc...
Good Luck po.
Check ko rin sa archives ko kung meron akong maidagdag. Maraming training pgn ako noon para sa chess engines na ginagawa kong opening books. Galing yon sa mga repertoire ng iba't ibang GM.
Mga isang linggo mong kapain yung Arena GUI at ChessX. Yung SCID/SCID vs. PC ay medyo mahirap sa newbies dahil di mo direktang ma-edit ang pgn. Yung Chessbase ang mas OK, though pag nasanay ka na sa mga format ng chess games ay maiitindihan mo rin.
Yung chessX at Arena GUI ang madali/magandang gamitin for chess analysis ng opening lines using Stockfish. Pwede mo ring i-annotate automatically pero di madaling gawin sa umpisa kung hindi mo pa kapa yung menus. Pero yung basic training tools ay mas preferred dahil direkta na sila by just following instructions. Kahit bata ay madaling matutong kalikutin yan.
Marami ring chess training sites sa net: google mo na lang at sangkaterba yon.
Mga isang linggo mong kapain yung Arena GUI at ChessX. Yung SCID/SCID vs. PC ay medyo mahirap sa newbies dahil di mo direktang ma-edit ang pgn. Yung Chessbase ang mas OK, though pag nasanay ka na sa mga format ng chess games ay maiitindihan mo rin.
Yung chessX at Arena GUI ang madali/magandang gamitin for chess analysis ng opening lines using Stockfish. Pwede mo ring i-annotate automatically pero di madaling gawin sa umpisa kung hindi mo pa kapa yung menus. Pero yung basic training tools ay mas preferred dahil direkta na sila by just following instructions. Kahit bata ay madaling matutong kalikutin yan.
Marami ring chess training sites sa net: google mo na lang at sangkaterba yon.
- Status