1:Transformer
- Zero-Based Analysis Tutorial [ Recommended ]
https://zhuanlan.zhihu.com/p/609271490
- Transformer detailed explanation [ recommended ]
https://wmathor.com/index.php/archives/1438/
- How to understand transformer from shallow to deep?
https://www.zhihu.com/question/471328838/answer/3011638037
2: The Prompt project is so powerful, do we still need model training?
https://www.zhihu.com/question/586331504/answer/2915043731
3: cs224n/slides courseware
https://web.stanford.edu/class/cs224n/slides/
4: Two lines of code enable LoRA fine-tuning && LLM emotional entity extraction practice
https://zhuanlan.zhihu.com/p/638247946
5: [QLoRA Practical Combat] Using a single card to efficiently fine-tune bloom-7b1, the effect is amazing
6: LLM training data construction example (deepspeedChat, chatGLM)
LLM training data construction example (deepspeedChat, chatGLM) - Zhihu
7: How to connect the local knowledge base to GPT?
https://www.zhihu.com/question/600738033/answer/3089765671
8: Why do current LLMs have Decoder only architecture?
https://www.zhihu.com/question/588325646/answer/2932131238
9: Prompt skill sharing: language learning assistants, knowledge cards and other knowledge content skills (common to all types of LLM)
10: If you want to learn large language models (LLM), which open source model should you start with?