잠시만 기다려 주세요.
JSCODE 박재성
2024. 9. 4.
🙋♂️ 도움이 필요하시거나 궁금한 점을 질문하고 싶을 땐 ? ✔️ https://open.kakao.com/o/gegroCTe ✔️ 취준생 & 현업 개발자 오픈채팅방 📙 비전공자도 이해할 수 있는 AWS 입문/실전 ✔️ https://inf.run/73afg ✔️ 20% 할인 쿠폰 코드 : 14206-1a845a617cc9 📗 프로그래밍 연합 동아리 합격 자소서 13종 ✔️ https://kmong.com/gig/377335 ✔️ 크몽 메시지로 유튜브 보고왔다고 말씀해주시면 20% 할인 해드리겠습니다😊 📓 상위 1% 개발자의 프로그래밍 과외 (멘토 : https://url.kr/z2n1ya ) ✔️프론트엔드 : https://kmong.com/gig/453361 ✔️백엔드(Spring 계열) : https://kmong.com/gig/485995 ✔️백엔드(Node 계열) : https://kmong.com/gig/452980 ✔️이력서/포폴 컨설팅 : https://kmong.com/gig/498515
[설명] 고가용성과 무중단배포 그리고 데이터베이스 [프로젝트해시태그] #server [챕터] 00:00 고가용성 06:02 롤링 배포 11:46 블루-그린 배포 13:42 카나리 배포 15:58 Master-Slave 18:10 CQRS
2024. 10. 13.
Developer Masterclass for $12 👉 https://dub.link/devcourse Thanks for watching! Please subscribe ❤️ Guillaume 👇 Read my adventures 👉 https://dub.link/newsletter Join my Discord 👉 https://dub.link/discord Book a coaching 👉 https://dub.link/guillaume Looking for a repository? 👉 https://dub.link/repos Visit my website 👉 https://dub.link/codewithg Install my package UserItem 👉 https://dub.link/useritem 💥 TOP VIDEOS Learn Nuxt 👉 https://dub.link/nuxt Learn Nextjs 👉 https://dub.link/next Learn Supabase 👉 https://dub.link/supabase ⚙️ • Theme that I use: VSCode Dark • VSCode Icons package: miguelsolorio.symbols • Snippets: my own, coming soon :) Thank you :) Guillaume
2024. 6. 29.
On this week's episode of the podcast, freeCodeCamp founder Quincy Larson interviews Emma Bostian. She's a software engineer turned manager at Spotify and Prolific coding teacher. We talk about: How at her first developer job at IBM, Emma's boss told her: "You need to get your stuff together or you won't make it in this industry." And the transformation that followed. Emma's thoughts on Computer Science degrees. "Going to college gives you credibility and a network. You can get opportunities that way." How Emma hires software engineers. (Hint: she tries to disregard degrees completely.) How Emma intentionally procrastinates some big tasks to give her mind time to figure out the puzzle pieces Can you guess what song I'm playing on my bass during the intro? It's from a 1979 punk song. Also, I want to thank the 10,776 kind people who support our charity each month, and who make this podcast possible. You can join them and support our mission at: https://www.freecodecamp.org/donate
2024. 8. 30.
LG AI 연구원이 EXAONE 3.0 7.8B 오픈소스 모델을 공개! - 7.8B 파라미터, 8T 학습 데이터 기반의 Decoder-only Transformer 모델 - 영어 성능: Real-world Use Cases 평균 점수 1위, Math와 Coding에서 특히 우수 - 한국어 성능: Real-world use cases와 단일 Benchmark 모두 평균 점수 1위 - 경제성: EXAONE 1.0 대비 비용 94% 절감, 2.0 대비 추론 시간 56% 감소 - 윤리적 투명성: Red Teaming 과정을 거쳐 평가, 결과 공개로 AI 윤리 발전 기대 - 국내외 AI 연구자들의 연구와 AI 생태계 발전에 기여 목적으로 공개 💥빵형의 수학없이 뽀개는 인공지능 입문 강의 ▶️신청링크 : https://bit.ly/3V1Jzmn ⭐️빵형의 차근차근 따라하는 수익형 웹서비스 강의 ▶️신청링크 : https://bit.ly/3UZkB83 💻소스코드 (엑사원 3.0): https://colab.research.google.com/drive/19VdXj-McHEAFH5yYDdyqColW9X2LK1Nk?usp=sharing EXAONE 3.0 7.8B on HuggingFace https://huggingface.co/LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct Chapters: 00:00 시작 01:25 본격적으로 질문공세 📮사업 및 개발문의 kairess87@gmail.com 😘빵형의 개발도상국 후원 카카오페이 : https://qr.kakaopay.com/Ej86nqvdu 투네이션 : https://toon.at/donate/helloworld 링크 모음 : https://lnk.bio/kairess #빵형 #엑사원 #exaone #ai #딥러닝 #chatgpt #챗gpt #lg
2024. 8. 10.
Fine-Tune Flux One: Create Hyper Realistic & Cartoon Images Easily! In this video, I demonstrate how to fine-tune Flux One, an advanced image generation model. Watch as I guide you through the process, from acquiring free credits on Fal.AI to uploading images, customizing settings, and generating your own images. Learn how to continually create consistent photos of various subjects, be it famous personalities, yourself, or animated characters. Explore the potential to use these images for a range of creative projects, such as children's books or videos. Dive in to see impressive results and practical applications of the fine-tuned model. https://fal.ai/models https://blackforestlabs.ai/ 00:00 Introduction to Fine-Tuning Flux One 00:17 Why Fine-Tune an Already Competent Model? 00:47 Getting Started with Foul AI 01:11 Uploading Images for Fine-Tuning 01:29 Adjusting Settings and Costs 02:21 Running Inference and Results 02:50 Potential Applications and Conclusion 03:03 Final Thoughts and Encouragement
2024. 8. 19.
저장소 링크: https://github.com/ejoongseok/k-wms
2023. 10. 1.