Update README.md
Browse files
README.md
CHANGED
@@ -8,6 +8,28 @@ This is a quantized gguf version of gemma-2-27b-it using an importance matrix (i
|
|
8 |
I hope it retains more Japanese support, but I can't be sure yet.
|
9 |
|
10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
Q4_K_Mをwiki.test.raw(英語)を使って計測したperplexityスコアが他の同等GGUF量子化モデルに比べて優れている事は確認済ですが理由はまだわかりません。
|
12 |
I have already confirmed that the perplexity score of Q4_K_M measured using wiki.test.raw is superior to other equivalent GGUF quantization models, but I don't know why yet.
|
13 |
|
@@ -21,6 +43,5 @@ There are unanswered questions.
|
|
21 |
- wiki.test.raw(英語)でperplexityを計測することにどこまで意味があるのか?
|
22 |
- (How far does it make sense to measure perplexity with wiki.test.raw (English)?)
|
23 |
|
24 |
-
Window 11のCMD/PowerShell
|
25 |
-
We have confirmed a problem that Japanese is garbled in CMD/PowerShell on Window 11.
|
26 |
-
We have confirmed that it works fine on Linux.
|
|
|
8 |
I hope it retains more Japanese support, but I can't be sure yet.
|
9 |
|
10 |
|
11 |
+
## 使い方(How to use)
|
12 |
+
|
13 |
+
[公式マニュアル](https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md)に従ってllama.cppをビルドします
|
14 |
+
Build llama.cpp according to the [official manual](https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md)
|
15 |
+
|
16 |
+
ダウンロードしたモデルを指定して下記コマンドを実行します
|
17 |
+
```
|
18 |
+
llama.cpp\build\bin\Release\llama-server -m .\gemma-2-27b-it-Q4_K_M.gguf
|
19 |
+
```
|
20 |
+
|
21 |
+
ブラウザで127.0.0.1を開きます
|
22 |
+
Open 127.0.0.1 in your browser
|
23 |
+
|
24 |
+
|
25 |
+
![image/png](https://cdn-uploads.huggingface.co/production/uploads/630469550907b9a115c91e62/9Yo4KEdh--5FCYjF46qWT.png)
|
26 |
+
|
27 |
+
![image/png](https://cdn-uploads.huggingface.co/production/uploads/630469550907b9a115c91e62/2eEhTIluRSLIPKrJCLsUw.png)
|
28 |
+
|
29 |
+
|
30 |
+
|
31 |
+
## その他の疑問など Other questions etc.
|
32 |
+
|
33 |
Q4_K_Mをwiki.test.raw(英語)を使って計測したperplexityスコアが他の同等GGUF量子化モデルに比べて優れている事は確認済ですが理由はまだわかりません。
|
34 |
I have already confirmed that the perplexity score of Q4_K_M measured using wiki.test.raw is superior to other equivalent GGUF quantization models, but I don't know why yet.
|
35 |
|
|
|
43 |
- wiki.test.raw(英語)でperplexityを計測することにどこまで意味があるのか?
|
44 |
- (How far does it make sense to measure perplexity with wiki.test.raw (English)?)
|
45 |
|
46 |
+
Window 11のCMD/PowerShellでは日本語が化けてしまう問題がある事を確認しています。
|
47 |
+
We have confirmed a problem that Japanese is garbled in CMD/PowerShell on Window 11.
|
|