Files
Hands-On-Large-Language-Models/bonus/3_quantization.md
2025-04-20 11:23:48 +02:00

17 lines
2.8 KiB
Markdown

<p align="center">
<a href="https://newsletter.maartengrootendorst.com/p/a-visual-guide-to-mixture-of-experts"><img src="../images/bonus_overview.png" width="100%" height="100%"></a>
</p>
# [A Visual Guide to Quantization](https://newsletter.maartengrootendorst.com/p/a-visual-guide-to-quantization) (Extends Chapters 7 and 12)
In the book, we cover the concept of quantization and showcase how it can be used to reduce computational requirements for both training and fine-tuning.
[A Visual Guide to Quantization](https://newsletter.maartengrootendorst.com/p/a-visual-guide-to-quantization) dives deeper into the technical intricacies of the technique, building upon the foundational concepts we introduced in the book.
<a href="https://newsletter.maartengrootendorst.com/p/a-visual-guide-to-quantization"><img src="../images/bonus_int8.png" width="50%" height="50%"></a>
The guide explore various quantization methods, such as as post-training quantization and quantization-aware training. It even showcases BitNet, a method by which we can get ternary parameters (3 values!).
<a href="https://newsletter.maartengrootendorst.com/p/a-visual-guide-to-quantization"><img src="../images/bonus_bitnet.png" width="50%" height="50%"></a>
By extending the book's coverage, readers can quickly get up to speed with the advanced topics covered in the guide. It serves as an excellent next step in the reading material.