File size: 873 Bytes
e2ba248
 
4ec1b85
 
d0d9041
 
e2ba248
4ec1b85
 
 
 
d0d9041
4ec1b85
 
 
 
d0d9041
4ec1b85
d0d9041
0e72b9d
d0d9041
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
---
license: cc-by-nc-4.0
language:
- ja
tags:
- gpt-neox
---
# matsuolab-weblab-10b-instruction-sft-gguf
[matsuo-labさんが公開しているweblab-10b-instruction-sft](https://huggingface.co/matsuo-lab/weblab-10b-instruction-sft)のggufフォーマット変換版です。

llama.cppのexamplesで動かせます。
*llama.cpp本家は開発速度が早くgptneoxのコンパイルが通らなくなる事もありますので、clone先をブランチに変更しました。*

## Usage (試用)

```
git clone --branch dev https://github.com/mmnga/llama.cpp.git
cd llama.cpp
make -j
./gptneox -m 'matsuolab-weblab-10b-instruction-sft-q4_0.gguf' -n 128 -t 8 -p '以下は、タスクを説明する指示です。要求を適切に満たす応答を書きなさい。\n\n### 指示:\n大規模言語モデルについて説明してください。\n\n### 応答:' 
```