If you want to understand what really happens under the hood of large language models - from pretrain to fine-tuning - Yandex had a cool intensive GPT Week.
I watched about half of it myself. Honestly, it's quite challenging in places. But there's a lot of really useful content, especially if you already have a background in ML and neural networks.
💡What they cover:
- How large language models are trained
- Stages of pretrain and fine-tuning
- What limitations and trade-offs are encountered in practice
📺 Recordings of all sessions can be found in the playlist.
And if you want not just to watch but to get hands-on, there are notebooks for the seminars:
- Seminar 1
- Seminar 2
- Seminar 4
There's also a digest of the intensive - brief and to the point.
⚡️ GPT Week is the next level after the "entry point". When you want to understand not only how to use models, but what really happens inside - from training to fine-tuning.
If some lectures seem difficult - that's normal. Even selective viewing gives a good idea of how engineering-complex LLMs are.
With this, the series on starting in ML and LLM logically concludes. I hope it helps you build your path faster and avoid unnecessary confusion 🙂
Back to table of contents
👩💻 Data Flow
Comments
0No comments yet.
Sign in to join the discussion.