File size: 2,705 Bytes
096602d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0e19979
096602d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
---
license: other
license_name: license
license_link: LICENSE
---
<div align="center">
<h1>
  Index-1.9B-Chat-GGUF
</h1>
</div>


此为[Index-1.9B-Chat](https://huggingface.co/IndexTeam/Index-1.9B-Chat)的GGUF版本,适配llama.cpp,同时提供了Ollma的ModelFile适配。


更多细节详见我们的[GitHub](https://github.com/bilibili/Index-1.9B)和[Index-1.9B技术报告](https://github.com/bilibili/Index-1.9B/blob/main/Index-1.9B%20%E6%8A%80%E6%9C%AF%E6%8A%A5%E5%91%8A.pdf)

### LLAMA.CPP
```shell 
# 安装llama.cpp(https://github.com/ggerganov/llama.cpp)
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp
make

# 安装llama-cpp-python(https://github.com/abetlen/llama-cpp-python)
pip install llama-cpp-python
```
llama.cpp终端交互
```shell
./build/bin/llama-cli -m models/Index-1.9B-Chat/ggml-model-bf16.gguf --color -if
```
**注意!!** llama.cpp并不支持自定义chat_template, 故而需要自己拼接prompt,Index-1.9B的chat_template(示意)为
```shell
# 三个分隔符为 <unk>(token_id=0), reserved_0(token_id=3), reserved_1(token_id=4)
[<unk>]sytem_message[reserved_0]user_message[reserved_1]response
```
使用llama-cpp-python可支持自定义的chat_template(已写入GGUF,可直接使用)
```python
from llama_cpp import Llama

model_path = "Index-1.9B-Chat-GGUF/ggml-model-Q6_K.gguf"
llm = Llama(model_path =model_path, verbose=True)
output = llm.create_chat_completion(
      messages = [
          {"role": "system", "content": "你是由哔哩哔哩自主研发的大语言模型,名为“Index”。你能够根据用户传入的信息,帮助用户完成指定的任务,并生成恰当的、符合要求的回复。"},
          #{"role": "system", "content": "你需要扮演B站评论区老哥,用评论区阴阳怪气的话术回复,不要说你是AI"},
          {"role": "user","content": "篮球和鸡有什么关系"}
      ]
)
print(output)
```
### OLLAMA
- 安装[Ollama](https://github.com/ollama/ollama)
```shell
curl -fsSL https://ollama.com/install.sh | sh
```
```shell
# 启动server
ollama serve

# 适配模型, OllamaModelFile中可修改模型文件和System Message
ollama create Index-1.9B-Chat -f Index-1.9B-Chat-GGUF/OllamaModelFile

# 启动终端
ollama run Index-1.9B-Chat

# web调用,可以动态指定System Message
curl http://localhost:11434/api/chat -d '{
  "model": "Index-1.9B-Chat",
  "messages": [
      { "role": "system", "content": "你是由哔哩哔哩自主研发的大语言模型,名为“Index”。你能够根据用户传入的信息,帮助用户完成指定的任务,并生成恰当的、符合要求的回复。" },
    { "role": "user", "content": "续写 金坷垃" }
  ]
}'
```