Posts
Page 6 - Showing 8 of 75 posts
View all posts by years →Decoding Strategies
Decoding strategies are the decision rules that turn a model’s next-token scores into actual tokens. The model provides a distribution over the vocabulary; the decoder decides how to pick the next token, how to stop, and how to trade off diversity vs. reliability.
1387 个字词
|
7 分钟
Collections
Tech Notes
Be a big fan of rising technologies.
Algorithm Notes
Woo. Time-tested wisdom.
强化学习合集
Notes on reading papers in the field of reinforcement learning.
Maths Laboratory
Do mathematics have a lab as well? Absolutely, with computer simulations.
生成式模型合集
Generative Models, modeling complex data distributions.
Tags

