Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

請加上 LZ 157 的 Po-25 #20

Open
farmer-lu opened this issue Jun 15, 2019 · 3 comments
Open

請加上 LZ 157 的 Po-25 #20

farmer-lu opened this issue Jun 15, 2019 · 3 comments

Comments

@farmer-lu
Copy link

farmer-lu commented Jun 15, 2019

小姐姐吉祥, 我是人工智慧行者 -- 君子風.

懇請加上 LZ 157 的 Po-25

/*

  1. 根據我的非正式統計, LZ 157 是計算效率最高的模型. (同等計算量之下的棋力最高)
  2. 並且是手機上最實用的模型. (因為在手機上 LZ 157 Po-25 每一手大約只花 2~3 秒, 而棋力應在六段以上)
  3. 後續我會寫一篇文章比較各模型, 並提交到本項目.
    */

謝謝小姐姐造福大家

@farmer-lu
Copy link
Author

farmer-lu commented Jun 15, 2019

如果配合 #19 , 那麼相鄰 Playouts 差四倍的 Po-1600, Po-400, Po-100, Po-25 都齊全了. 形成完美的等比數列, 可以觀察出 Playouts 和 棋力的關係, 也是一大發現.

@farmer-lu
Copy link
Author

farmer-lu commented Jun 16, 2019

模型的效率分析已初步寫到 這裡 :

https://github.com/farmer-lu/computer-go/blob/master/model_efficiency_zh.md

Repository owner deleted a comment Jul 4, 2019
@nanzi
Copy link

nanzi commented Oct 11, 2019

Low playouts make a fuzz on AI strength.
You can try new 15b192f weight,the best for now is 15b-238-136k by bjiyxo from LeelaZero official website.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants