dahara1 commited on
Commit
26f621c
·
verified ·
1 Parent(s): 0867a86

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -127,6 +127,7 @@ The following information may be helpful in your further exploration.
127
  以下の情報は更なる探求を行う際に参考になるかもしれません。
128
  - [About imatrix overfitting, and importance of input text](https://github.com/ggerganov/llama.cpp/discussions/5263)
129
  - [Importance matrix calculations work best on near-random data](https://github.com/ggerganov/llama.cpp/discussions/5006)
 
130
  - [llama.cpp:iMatrix量子化は日本語性能にどう影響するか?](https://sc-bakushu.hatenablog.com/entry/2024/04/20/050213)
131
  - [Command R+はどこまで量子化するとアホになってしまうのか?](https://soysoftware.sakura.ne.jp/archives/3834)
132
  - [GGUFって結局どのサイズ選んだらいいの??](https://zenn.dev/yuki127/articles/e3337c176d27f2)
 
127
  以下の情報は更なる探求を行う際に参考になるかもしれません。
128
  - [About imatrix overfitting, and importance of input text](https://github.com/ggerganov/llama.cpp/discussions/5263)
129
  - [Importance matrix calculations work best on near-random data](https://github.com/ggerganov/llama.cpp/discussions/5006)
130
+ - [Synthetic nonsense data improves llama.cpp Quantization accuracy](https://www.reddit.com/r/LocalLLaMA/comments/1ah3w8d/comment/kouw5aj/?context=3)
131
  - [llama.cpp:iMatrix量子化は日本語性能にどう影響するか?](https://sc-bakushu.hatenablog.com/entry/2024/04/20/050213)
132
  - [Command R+はどこまで量子化するとアホになってしまうのか?](https://soysoftware.sakura.ne.jp/archives/3834)
133
  - [GGUFって結局どのサイズ選んだらいいの??](https://zenn.dev/yuki127/articles/e3337c176d27f2)