일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | ||||||
2 | 3 | 4 | 5 | 6 | 7 | 8 |
9 | 10 | 11 | 12 | 13 | 14 | 15 |
16 | 17 | 18 | 19 | 20 | 21 | 22 |
23 | 24 | 25 | 26 | 27 | 28 |
Tags
- 자연어처리
- GenAI
- gcp
- 코드업
- Microsoft
- LeetCode
- Python3
- codeup
- Python
- 생성형AI
- C#
- 데이터사이언스
- 클라우드
- Blazor
- 리트코드
- Azure
- 릿코드
- 파이썬기초100제
- TwoPointer
- 코드업파이썬
- 파이썬
- nlp
- 머신러닝
- 파이썬기초
- 투포인터
- GenerativeAI
- 알고리즘
- 파이썬알고리즘
- 구글퀵랩
- 빅데이터
Archives
- Today
- Total
Tech for good
[Google Cloud Skills Boost(Qwiklabs)] Introduction to Generative AI Learning Path - 2. Introduction to Large Language Models 본문
IT/Cloud
[Google Cloud Skills Boost(Qwiklabs)] Introduction to Generative AI Learning Path - 2. Introduction to Large Language Models
Diana Kang 2023. 8. 27. 16:18https://www.youtube.com/playlist?list=PLIivdWyY5sqIlLF9JHbyiqzZbib9pFt4x
Generative AI Learning Path
https://goo.gle/LearnGenAI
www.youtube.com
- Pre-trained: for general purpose with a large data set
- Fine-tuned: for specific aims with a much smaller data set
- Transformer model (e.g. PaLM)
- A transformer model consists of encoder and decoder.
- Encoder - encodes the input sequence and passes it to the decoder
- Decoder - decodes the representations for a relevant task
- Traditional programming -> 고양이를 구별하기 위해 규칙들을 하드코딩
- 신경망 기반 접근법 -> 고양이와 개 사진을 미리 주고, 예측하게 함.
- 생성형 접근법 -> generate our own content (e.g. text, images, audio, video, etc..)
- 생성형 AI 모델들: ingest very large data from multiple sources across the Internet and build foundation language models