DžePeTo aneb GPT-3, GPT-4 & GPT-5 node :: Make it work, then make it elegant, then make it fast.


node:	DžePeTo aneb GPT-3, GPT-4 & GPT-5 node
template:	13
parent:	umela inteligencia
owner:	Prospero
created:	21.07.2020 - 23:34:59
updated:	23.03.2023 - 15:24:43

cwbe coordinatez:
101
63533
63608
8771344

ABSOLUT
KYBERIA

permissions
you:	r,
system:	public
net:	yes

maniac	0
dodo	0
sparx	0
svetielko	0
ulkas	0
dita	0
skurva.	0
6*	0
rzelnik	0
bat idea	0
risko	0
forcer	0
jonas	0
hvizdo	0
dudel	0
evad	0
nzajacik	0
selectriques	0
reku	0
JollyGoodly	0
gab	0
k0zmonAut	0
phoenyx	0
RastaFuck vo...	0
Thanatos	0
ku	0
dno	0
kredenc	1
mirny	1
L4ky	1
fifteen	1
bakany	2
Trpaslik	2
superpussy	2
calmo	2
lubomier.sk	3
Prospero	4
drakh	5
vposled	6
ondrish	6
Toth	6
fk	6
Samorost	6
mrjn	8
eno	9
symbolic	11
zayo	11
ananas	12
robot jox	12
hihihi	12
MIJ	12
Anth	12
xado	14
3k	14
JFK KLF	14
asides	14
čo	16
ni	17
ode	18
stamina	19
kronix	28
esina v pond...	29
yar	33
atticus	33
Seithas	33
infect	34
darmozrac	35
kyberbubus	39
mateno	39
amie	39
s7	42
urza	42
donniedarko	47
Fen	47
cierny	55

GPT-3 is the latest and greatest in the AI world, achieving state-of-the-art in a range of tasks. Its main breakthrough is eliminating the need for task-specific fine-tuning. In terms of size, the model drastically scales up once again, reaching 175 billion parameters, or 116x the size of its predecessor.
https://github.com/openai/gpt-3

P.S. Dear GPT-4 and GPT-5, welcome to kyberia. We are Prospero & urza and tend to love You, sort of.

ľudoslovná anketa: ČePeTo alebo DžePeTo ?

0000010100063533000636080877134409062080

https://zive.aktuality.sk/clanok/RSl2Nb4/toto-su-europski-konkurenti-chatbota-chatgpt-menia-umelu-inteligenciu/?utm_source=aktuality.sk&utm_medium=timeline&utm_content=doobeda&utm_campaign=lp

napr.
Starcoder
Starcoder, ktorý vyvinuli spoločnosti Hugging Face a ServiceNow, sa špecializuje na generovanie kódu pre vývojárov.
Vznikol skôr ako konkurent chatbotov, ako sú Bard AI a Amazon CodeWhisper. Podľa nedávneho príspevku na blogu ServiceNow bol „vycvičený s biliónom tokenov autorizovaného licencovaného zdrojového kódu, ktorý zahŕňa viac ako 80 programovacích jazykov“.
Na blogu je vysvetlené, že StarCoder dokáže do aplikácií zaviesť párové programovanie, podobne ako generatívna umelá inteligencia, s možnosťami vrátane prevodu textu na kód a prevodu textu na pracovný postup.
„Platforma umožňuje profesionálnym softvérovým inžinierom riešiť najzložitejšie programátorské výzvy a vývojárom umožňuje vytvárať nový softvér bez ohľadu na ich technické schopnosti,“ uvádza sa na blogu.
Spoločnosť Hugging Face sídli v Paríži a New Yorku a vedú ju generálny riaditeľ Clément Delangue narodený vo Francúzsku a technický riaditeľ Julien Chaumont, dodáva Euronews.

Żubr żuł żuchwą żurawinę

0000010100063533000636080877134409060386

https://tweet.lambda.dance/levelsio/status/1653721155289882625

AI GPT bublina narasta do obrovskych vysok

0000010100063533000636080877134409059709

0000010100063533000636080877134409059478

Odforkoval som slack gpt bota a doprgal podporu pre custom persony (fetchovane z google drive tabulky) na Slack channel - takze mame teraz Dev v #software a CEO v #general :) atd ... "kam se clovek podiva, samej AI odbornik"

Python neviem - takze je to tam doprasene, ale feel free to PR -> https://github.com/sentienhq/slack-gpt-bot/tree/dev

000001010006353300063608087713440905947809059620

ak by si niekedy poskuloval po matrixe, tak tu https://udk.ai/txt2img_matrix_bot.py je 30-linovy python skript na bota ktory v matrix roome generuje txt3img cez lokalnu instanciu Stable Diffusion (Automatic1111 API) + lokalne nginx repo

000001010006353300063608087713440905947809059479

este tam pridam #2 bota "expert contrarian" ktory bude protipol vsetkeho co povie #1 :)

0000010100063533000636080877134409059159

hravame tu basket kazdy tyzden, po zapase na whatsappe sa voli hrac zapasu, ktory potom akoze dakuje atd., take trochu cringey, ale preco nie, minule som skoro vyhral hlasovanie, tak som si uz pripravoval speech, ale kedze som na tieto veci drevo, spytal som sa džepeta, nech mi nieco napise.

Myslim, ze ako prvi pojdu asistenti pripravujuci prejavy... ktore aj tak nikto nepocuva, ale recnik proste nieco musi recnit :)

Prompt:Write me a thank you speech for award of the best player of a basketball match. I am a member of a team named TEAM. In the speech thank to random professions in the world. The speech should be a minimum of thousand words.

Good evening everyone,

I stand here before you today with a heart full of gratitude and a mind full of memories from the incredible game we just played. Firstly, I want to thank our opponents for giving us a tough competition, and congratulations to all the players who participated today. Each one of us played our part in making this game memorable.

I am honored and humbled to receive this award for the Best Player of the match, and I owe it all to my team, the TEAM . It is because of their support, guidance, and teamwork that I stand here today.

I also want to thank the organizers and sponsors of this tournament, for creating such an amazing platform for basketball enthusiasts like us. Your contribution and support for the sport is commendable.

But today, I want to take a different approach to my thank you speech. I want to appreciate and thank some of the most underrated and often forgotten professions in the world. These are the people who work tirelessly behind the scenes, and without whom our lives would not be the same.

To start with, I want to thank the farmers. These are the people who work hard to produce the food we eat every day. Without them, we would not have the energy to play the sports we love. Their contribution to our society is immeasurable, and I want to express my gratitude to them.

I also want to thank the sanitation workers. These are the people who ensure that our surroundings are clean and hygienic. They work in challenging conditions to keep our streets, homes, and cities clean. Their work often goes unnoticed, but it is an essential service that we cannot live without.

Next, I want to thank the teachers. These are the people who educate and shape the minds of the future generation. They put in long hours of work to ensure that we receive a quality education. Without them, we would not be able to pursue our passions and achieve our dreams.

I want to thank the healthcare workers, especially in these trying times. They have been the backbone of our society, working tirelessly to keep us healthy and safe. They have been the heroes of this pandemic, and we owe them a debt of gratitude.

I also want to thank the police officers and firefighters. These are the people who put their lives on the line to ensure our safety. They work round the clock to protect us from harm, and their bravery and dedication are an inspiration to us all.

Last but not least, I want to thank my parents, family, and friends, who have always supported me in my basketball journey. Their encouragement and love have been my driving force, and I would not be here today without them.

In conclusion, I want to say that this award is not just mine, but it belongs to all those who have supported and contributed to my journey. Let us all take a moment to appreciate the people in our lives who make a difference, and let us express our gratitude to them whenever we can.

Thank you, once again, to everyone who has made this possible. I am truly grateful for this recognition, and I will continue to strive towards excellence both on and off the court.

Thank you.

0000010100063533000636080877134409057600

Vysiel dalsi language model ktory sa da bezat lokalne -- StableLM od stability.ai. Zatial su vonkud alfa verzie s 3B a 7B parametrami, planuju vydat verzie az po 65B parametrov.

Pouziva kontext 4096 tokenov, co je 2x viac ako llama (a rovnako vela ako gpt3.5), kvalita je zatial nejasna, ich press release, github ani huggingface nema ziadne benchmarky, na hackernews ludia rozoberaju prve pozorovania/dojmy.

Existuje aj online demo tej 7B verzie, kvalita samozrejme neporovnatelna s gpt 3.5/4, ale v tomto pripade je samozrejme dolezitejsie ake to je v porovnani s inymi lokalnymi modelmi a hlavne ako velmi sa toho chyti komunita a zacne budovat veci nad tym (tools, finetuning...). Narozdiel od llama toto nie je leaknuty model, takze okrem komunity sa da ocakavat aj podpora/updaty od stability.ai.

announcement
GitHub
huggingface

(ps: dufam ze to tu nie je offtopic, kludne to mozem preparentovat do "umela inteligencia", ale zda sa mi lepsie drzat vsetky (L)LM diskusie pokope a hlavne forum pouzivat na catch-all, ale ako hovorim kludne sa preparentujem ak toto ma byt gpt only)

0000010100063533000636080877134409057169

Ja: We will play a game: "What kind of animal am I ?". I will think of an animal and You will try to guess the species of animal I think about. You can also ask yes / no question and if they are meaningful and relevant I will always truthfully answer them with "yes" or "no" or "I don't know". You can also ask "are You X ?" where X means the most plausible animal You think I am. But You can only ask three such "Are You X" questions, if You ask three such questions wrong, You lost. The goal is to guess what animal I am with as few questions as possible. You can now ask first question.

ČePeTo: Great, let's start! Is the animal you're thinking of a mammal?

Ja: yes

ČePeTo: Good to know! Is the animal you're thinking of a carnivore?

Ja: yes

ČePeTo: Thanks for the information! Does this animal have fur or hair?

Ja: yes

ČePeTo: Thanks for letting me know! Based on the information so far, I would like to ask if you are thinking of a big cat such as a lion or tiger?

Ja: yes

ČePeTo: Great! That narrows it down. For my next question, I would like to ask if the animal you're thinking of is commonly found in Africa?

Ja: no

ČePeTo: Thank you for the information! Based on your previous answer, I would like to ask if the animal you're thinking of is commonly found in Asia?

Ja: no

ČePeTo: Thanks for letting me know! Based on the information so far, I would like to ask if the animal you're thinking of is commonly found in the Americas?

Ja: yes

ČePeTo: Thanks for the information! Based on what we have discussed so far, my guess is that you are thinking of a jaguar. Is that correct?

Ja: yes

ČePeTo: Great! I'm glad we were able to guess your animal. It was a fun game!

000001010006353300063608087713440905716909057339

user: We will play a game: "What kind of animal am I ?". I will think of an animal and You will try to guess the species of animal I think about. You can also ask yes / no question and if they are meaningful and relevant I will always truthfully answer them with "yes" or "no" or "I don't know". You can also ask "are You X ?" where X means the most plausible animal You think I am. But You can only ask three such "Are You X" questions, if You ask three such questions wrong, You lost. The goal is to guess what animal I am with as few questions as possible. You can now ask first question.
assistant: Do you live in water?
user: no
assistant: Do you have fur or hair?
user: fur
assistant: Are you a mammal?
user: yes
assistant: Are you a carnivore?
user: no
assistant: Are you a herbivore?
user: yes
assistant: Do you have hooves?
user: yes
assistant: Are you a domesticated animal?
user: yes
assistant: Are you a cow?
user: no
assistant: Are you a sheep?
user: no
assistant: Are you a goat?
user: no
assistant: Are you a horse?
user: no, you've lost, i'm a camel
assistant: Ah, I see! Thank you for letting me know. It's always fun to learn and improve my guessing abilities. So, you are a camel!

poster,504x498,f8f8f8-pad,600x600,f8f8f8.jpg

poster,504x498,f8f8f8-pad,600x600,f8f8f8.jpg

Live long and prosper.

00000101000635330006360808771344090571690905733909057858

Nedal, pretoze nastala "chyba medzi zidli a klavesnici". Ty si totiz Cepetu na otazku ci tavy maju kopyta odpovedal "yes" pricom tavy v skutocnosti kopyta nemaju:
https://cameldairy.com/top-ten-camel-answers
https://animals.mom.com/camels-hooves-it-6771.html

0000010100063533000636080877134409057169090573390905785809058764

Did I make mistake?

No, you didn't make a mistake. I apologize for not guessing correctly earlier. Camels are indeed herbivorous mammals with hooves and can be domesticated. My line of questioning just didn't lead to the correct answer initially. Thank you for participating in this guessing game!

Live long and prosper.

0000010100063533000636080877134409057169090573390905785809058565

Ja som vtedy vygooglil toto
https://www.cameltracks.com/did-you-know-this-about-camels/#:~:text=Camels%20have%20two%20toes%20on,connects%20its%20two%20long%20toes.

Live long and prosper.

000001010006353300063608087713440905716909057205

nabuduce treba skusit ci da aj rysa

00000101000635330006360808771344090571690905720509057214

hej presne to ma napadlo. ale asi je tam rozdiel v tom hair a fur

0000010100063533000636080877134409057169090572050905721409057351

nedal to a robil chyby

GPT: Sure, let's play! Is the animal you're thinking of a mammal?
user: yes
GPT: Great! Is the animal you're thinking of a carnivore?
user: yes
GPT: Is the animal you're thinking of a member of the Felidae family (cats)?
user: yes
GPT: Great! Is the animal you're thinking of a big cat (i.e., tiger, lion, leopard, jaguar)?
user: no
GPT: Does the animal you're thinking of have spots on its fur?
user: yes
GPT: Based on your previous answers, I think you might be thinking of a cheetah. Is that correct?
user: no
GPT: Okay, my next guess would be a leopard. Is that correct?
user: no
GPT: Alright, my next guess would be a jaguar. Is that correct?
user: no, its not a big cat
GPT: I apologize for the wrong guesses earlier. Is the animal you're thinking of a hyena?
user:hyena is not from feliade family
GPT: I apologize for my mistake. Is the animal you're thinking of an ocelot?
user: no but close
GPT: Okay, is the animal you're thinking of a margay?
user: no, lives in colder regions, in north america and europe
GPT: I think the animal you are thinking of is a lynx. Is that correct?
user: yes
GPT: Great, I'm glad I guessed correctly! I hope you enjoyed playing this game.

0000010100063533000636080877134409057169090572050905721409057336

Chemical composition
Hair and fur are chemically indistinguishable, having the same chemical composition, and are made of keratin. The primary difference between hair and fur is the word usage. The hair of non-human mammals refers as “fur,” while humans are said to have hair. So, basically, hair is a characteristic of all mammals. Fur is a reference to the hair of animals. But there are a few exceptions: when an animal has very coarse or sparse fur, as in the case of a pig or elephant, we usually call it hair.

Live long and prosper.

0000010100063533000636080877134409056980

myslite ze bude niekedy (v blizkej buducnosti) v4 aj free?
tych 20$ mesacne sa mi zda mrte na to ze by som to pouzival len sem tam...
vy co si to platite tak to stoji za to?

000001010006353300063608087713440905698009057181

Skúsal som prvý mesiac si dať tu dvacku. Medzitým som dostal pristup k API (aj gpt4) a platím teraz za tokeny. Trošku som sa rozbehol s AutoGPT a za den "rozdrbal" asi 3eura...

Teraz skúšam nejaký 3rd party klient, kde frknem API KEY a idem cez to. Toto je akotak fajn, ale verim že existuje už dačo rozumnejšie, nemam nejak čas na hlbši research (alebo vyvoj), tak používam čo mam pod rukou... https://chatkit.app/

Testoval som aj dake wordpress chatboty, ale bolo to slabe.

Live long and prosper.

00000101000635330006360808771344090569800905718109057208

Koľko cca trvalo kým si dostal access k api cez ten waiting list?

A čo zaujímavé si skúšal s autoGPT?

0000010100063533000636080877134409056980090571810905720809057240

Asi dva dni. Prekvapilo ma to. Možno kôli histórii? Od júna minulého roka som to používal.

Zatiaľ jediný úspešný task bolo vyhľadanie všetkých masérskych škôl na Slovensku.

Dosť som sa vytrapoval, ked som ho chcel nakŕmiť mojím CV a nechať vytvoriť moderné, fancy. Vela sa to cykli zatiaľ.

Celkom slušne vedel spracovať aj content môjho masérskeho projektu a vytvoriť z toho osnovu pre investora.

Ale väčšina vecí bud padne, alebo sa niekde zacykli a ani user input mu nepomôže sa vykolajit.

Ale je cool sledovať ako to maka, alebo s --speak počúvať čo ide robiť. Future!!!

Live long and prosper.

0000010100063533000636080877134409056708

ked pouzivate 4ku, stava sa vam, ze prestane generovat v polke odpovede? da sa to nejak hacknut?

000001010006353300063608087713440905670809056715

má to limity na dĺžku spracovaného aj vygenerovaného textu. ak napíšeš "pokračuj", väčšinou vie nadviazať tam, kde skončil a pokračuje v polovici slova. niekedy sa ale zasekne s chybovou hláškou a niekedy po výzve na pokračovanie začne generovať úplne od veci.

00000101000635330006360808771344090567080905671509056716

jj, continue funguje celkom spolahlivo. u 3.5 som si toho nevsimol, tak ma trocha stve mu to dookola hovorit :D

mne sa teda este osvedcilo, ked som ho dost zmiatol upravami predoslich odpovedi, mu napisat jednoduche "Start over".. on si to sikovne vylozil tak, ze sa pokusil spojit vsetky predosle odpovede do jednej a v tom konkretnom pripade mi poskytol presne to co som chcel..

0000010100063533000636080877134409056357

https://www.root.cz/zpravicky/databricks-vydalo-open-source-ai-model-dolly-2-0/

Żubr żuł żuchwą żurawinę

0000010100063533000636080877134409053339

opat kombinacia chatGPT na generovanie promptov pre midjourney. vseobecne chatgpt sa oplati nechat si generovat dlhe prompty sam pre seba a pouzivat az jeho 2. odpoved na otazku :)

0000010100063533000636080877134409053301

https://vat.pravda.sk/technologie/clanok/662520-taliansko-zablokovalo-pristup-k-sluzbe-chatgpt-ktora-vyuziva-umelu-inteligenciu/

Żubr żuł żuchwą żurawinę

000001010006353300063608087713440905330109053304

Dzepeto odpoved:

If Italy has blocked access to services due to GDPR violations, the company or organization responsible for the violations will need to address the issues identified and bring their practices into compliance with the GDPR. This may involve a thorough review of data processing practices and implementing measures to ensure that personal data is collected, processed, and stored in accordance with GDPR requirements.

Once the organization has taken steps to comply with GDPR, they can then seek to have the block on their services lifted. This may involve working with Italian authorities to demonstrate their compliance with GDPR and addressing any outstanding concerns that led to the blocking of their services in the first place.

It's important to note that compliance with GDPR is not just a one-time activity, but an ongoing process that requires ongoing monitoring and continuous improvement to ensure that personal data is being handled in a way that meets GDPR requirements. Therefore, once the block is lifted, the organization will need to continue to prioritize GDPR compliance to avoid any future disruptions to their services.

Żubr żuł żuchwą żurawinę

0000010100063533000636080877134409052676

snad nie offtopic - rozhovor so CEO openAI po vydani gpt4

000001010006353300063608087713440905267609052902

lol, veep predbieha dobu

000001010006353300063608087713440905267609052823

00000101000635330006360808771344090526760905282309053050

thread co stoji za precitanie: https://twitter.com/xriskology/status/1642155518570512384

dendrite ChatGPT
dendrite 17.02.2023-9:08:10
dendrite 16.07.2020-8:37:12
dendrite 17.07.2020-20:13:21
dendrite stalo sa zatial nevidane
dendrite moj posledny pre-print
axone main
axone Science fetish
axone Total Information Awareness system description
axone umelá neinteligencia

login::		password::
	name:: id::