{"id":2829,"date":"2023-06-22T00:16:11","date_gmt":"2023-06-21T15:16:11","guid":{"rendered":"https:\/\/brand-master.net\/?p=2829"},"modified":"2023-06-22T00:16:11","modified_gmt":"2023-06-21T15:16:11","slug":"entendendo-como-chatgpt-funciona-rodando-sua-propria-ia","status":"publish","type":"post","link":"https:\/\/brand-master.net\/?p=2829","title":{"rendered":"Entendendo Como ChatGPT Funciona - Rodando sua Pr\u00f3pria IA"},"content":{"rendered":"<div class=\"youtuber-press-title\">Entendendo Como ChatGPT Funciona - Rodando sua Pr\u00f3pria IA<\/div>\n<div class=\"youtuber-press\">\n    <iframe loading=\"lazy\" width=\"560\" height=\"315\" src=\"https:\/\/www.youtube.com\/embed\/O68y0yRZL1Y\" frameborder=\"0\" allowfullscreen=\"allowfullscreen\"><\/iframe>\n<\/div>\n<div class=\"youtuber-press-thumbnail-wrapper\" style=\"margin-top: 1rem\">\n    <img decoding=\"async\" class=\"youtuber-press-thumbnail\" src=\"https:\/\/i.ytimg.com\/vi\/O68y0yRZL1Y\/hqdefault.jpg\">\n<\/div>\n<div class=\"youtuber-press-description\">\nN\u00e3o, AGI n\u00e3o est\u00e1 pr\u00f3ximo de acontecer. <\/p>\n<p>Quais s\u00e3o os limites da atual tecnologia de transformadores generativos pr\u00e9-treinados, vulgo GPT?<br \/>\nO que s\u00e3o os tais \"modelos\" que eles usam?<br \/>\nO que s\u00e3o os tais \"par\u00e2metros\".<br \/>\nPorque as IAs parecem humanas nas respostas?<br \/>\nIAs v\u00e3o substituir programadores??<br \/>\nComo \u00e9 poss\u00edvel conseguir rodar um clone de ChatGPT na sua pr\u00f3pria m\u00e1quina, totalmente offline? <\/p>\n<p>Vamos responder tudo isso e muito mais.<\/p>\n<p>## Cap\u00edtulos<\/p>\n<p>00:01:07 - CAP 01 - Minha IA PESSOAL e OFFLINE - text-generation-webui<br \/>\n00:06:37 - CAP 02 - Introdu\u00e7\u00e3o a MODELOS - Como IA Aprende?<br \/>\n00:10:09 - CAP 03 - Auto-Corretor de Teclado de Celular - Cadeias de Markov<br \/>\n00:15:34 - CAP 04 - 170 TRILH\u00d5ES de par\u00e2metros - Maior que nosso c\u00e9rebro?<br \/>\n00:20:28 - CAP 05 - De Markov a IA - O que s\u00e3o Tensors?<br \/>\n00:30:53 - CAP 06 - A Revolu\u00e7\u00e3o da Auto-Aten\u00e7\u00e3o - O que s\u00e3o Transformers?<br \/>\n00:35:24 - CAP 07 - Hardware pra Treinar IA - NVIDIA<br \/>\n00:42:09 - CAP 08 - Par\u00e2metros e Sinapses s\u00e3o equivalentes? - Quantiza\u00e7\u00e3o<br \/>\n00:49:42 - CAP 09 - Estamos pr\u00f3ximos de AGI? IAs s\u00e3o Inteligentes?<br \/>\n00:52:10 - CAP 10 - Controlando Criatividade de IAs - Temperatura<br \/>\n01:00:27 - CAP 11 - A Av\u00f3 do ChatGPT - Eliza<br \/>\n01:04:29 - CAP 12 - IAs que fazem Imagens - Diffusion<br \/>\n01:09:31 - CAP 13 - Discutindo PROMPTs - LangChain<br \/>\n01:13:08 - CAP 14 - IAs v\u00e3o matar programadores? Quais os Limites?<br \/>\n01:22:15 - CAP 15 - Virando \"Programador 10x\" com aux\u00edlio de IAs - Copilot<br \/>\n01:23:47 - CAP 16 - CONCLUS\u00c3O - O come\u00e7o da BOLHA<br \/>\n01:28:18 - Bloopers<\/p>\n<p>## Links<\/p>\n<p>* [Introducing LLaMA: A foundational, 65-billion-parameter language model (facebook.com)](<br \/>\n* [NVIDIA\/Megatron-LM](<br \/>\n* [oobabooga\/text-generation-webui](<br \/>\n* [TheBloke\/Wizard-Vicuna-30B-Uncensored-GPTQ](<br \/>\n* [Neural networks - YouTube](<br \/>\n* [GPT-4 Will Have 100 Trillion Parameters \u2014 500x the Size of GPT-3 | by Alberto Romero | Towards Data Science](<br \/>\n* [MIT Deep Learning 6.S191 (introtodeeplearning.com)](<br \/>\n* [DF Direct Special: Starfield Tech Breakdown - 30FPS, Visuals, Rendering Tech + Game Impressions - YouTube](<br \/>\n* [NVIDIA Made a CPU.. I\u2019m Holding It. - YouTube](<br \/>\n* [(2) Artem Andreenko \ud83c\uddfa\ud83c\udde6 on Twitter: \"I've sucefully runned LLaMA 7B model on my 4GB RAM Raspberry Pi 4. It's super slow about 10sec\/token. But it looks we can run powerful cognitive pipelines on a cheap hardware. \" \/ Twitter](<br \/>\n* [\u00c9mile Borel and the Infinite Monkey Problem | SciHi Blog](<br \/>\n* [GitHub - codeanticode\/eliza: The classic Eliza psychologist program, implemented as a Processing library.](<br \/>\n* [Stories \u2022 Instagram](<br \/>\n* [TheBloke (Tom Jobbins) (huggingface.co)](<br \/>\n* [Introduction | \ud83e\udd9c\ufe0f\ud83d\udd17 Langchain](<\/p>\n<p>* Podcast:<br \/>\n* Transcript:\n<\/p><\/div>\n","protected":false},"excerpt":{"rendered":"<p>Entendendo Como ChatGPT Funciona - Rodando sua Pr\u00f3pria IA N\u00e3o, AGI n\u00e3o est\u00e1 pr\u00f3ximo de acontecer. Quais s\u00e3o os\u2026<\/p>\n","protected":false},"author":1,"featured_media":2830,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[43],"tags":[44],"class_list":["post-2829","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-chatgpt","tag-chatgpt"],"_links":{"self":[{"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/posts\/2829","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/brand-master.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2829"}],"version-history":[{"count":0,"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/posts\/2829\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/brand-master.net\/index.php?rest_route=\/wp\/v2\/media\/2830"}],"wp:attachment":[{"href":"https:\/\/brand-master.net\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2829"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/brand-master.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2829"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/brand-master.net\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2829"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}