# openai-simple-rerank
一个简单的rank模型的OPENAI兼容API服务的客户端访问工具。
## 安装
```shell
pip install openai-simple-rerank
```
## 使用
### 建议使用xinference启动rerank模型
*验证过的模型列表*
- bge-reranker-v2-m3
### 配置变量设置
```shell
# OPENAI兼容API服务，可以xinference提供
export OPENAI_BASE_URL="http://localhost/v1"
# OPENAI兼容API服务密钥，一般以sk-开头，共16位长
export OPENAI_API_KEY=""
# 默认的文本重排序模型
export OPENAI_RERANK_MODEL="bge-reranker-v2-m3"
# 默认的最大字符串长度控制
export OPENAI_RERANK_MAX_SIZE=1024
```
### 获取文本向量
*代码*
```python
from openai_simple_rerank.base import get_rerank_scores
s1, s2 = get_rerank_scores("hello", ["hi", "world"])
print(s1, s2)
```
*输出*
```txt
0.9953891038894653 0.40583446621894836
```
## 关于字符串长度控制
- 模型本身一般没有字符串长度控制。
- 但过长的字符串会导入模型占用内存的增长。
- 所以默认将字符串长度控制在：1024字。
- 通过`OPENAI_RERANK_MAX_SIZE`设置默认最大字符串长度。
- 也可以函数调用中指定最大字符串长度。
- 注意：所有超过最大长度的字符串将被截断。
## 版本记录
### v0.1.0
- 版本首发。
Raw data
{
"_id": null,
"home_page": null,
"name": "openai-simple-rerank",
"maintainer": "rRR0VrFP",
"docs_url": null,
"requires_python": null,
"maintainer_email": null,
"keywords": "openai-simple-rerank",
"author": "rRR0VrFP",
"author_email": null,
"download_url": "https://files.pythonhosted.org/packages/85/b5/a53528ec4faac0faf0beca381fd85f86cdccd7050db1ea238417b52526b0/openai-simple-rerank-0.1.0.tar.gz",
"platform": null,
"description": "# openai-simple-rerank\r\n\r\n\u4e00\u4e2a\u7b80\u5355\u7684rank\u6a21\u578b\u7684OPENAI\u517c\u5bb9API\u670d\u52a1\u7684\u5ba2\u6237\u7aef\u8bbf\u95ee\u5de5\u5177\u3002\r\n\r\n## \u5b89\u88c5\r\n\r\n```shell\r\npip install openai-simple-rerank\r\n```\r\n\r\n## \u4f7f\u7528\r\n\r\n### \u5efa\u8bae\u4f7f\u7528xinference\u542f\u52a8rerank\u6a21\u578b\r\n\r\n*\u9a8c\u8bc1\u8fc7\u7684\u6a21\u578b\u5217\u8868*\r\n\r\n- bge-reranker-v2-m3\r\n\r\n### \u914d\u7f6e\u53d8\u91cf\u8bbe\u7f6e\r\n\r\n```shell\r\n# OPENAI\u517c\u5bb9API\u670d\u52a1\uff0c\u53ef\u4ee5xinference\u63d0\u4f9b\r\nexport OPENAI_BASE_URL=\"http://localhost/v1\"\r\n# OPENAI\u517c\u5bb9API\u670d\u52a1\u5bc6\u94a5\uff0c\u4e00\u822c\u4ee5sk-\u5f00\u5934\uff0c\u517116\u4f4d\u957f\r\nexport OPENAI_API_KEY=\"\"\r\n# \u9ed8\u8ba4\u7684\u6587\u672c\u91cd\u6392\u5e8f\u6a21\u578b\r\nexport OPENAI_RERANK_MODEL=\"bge-reranker-v2-m3\"\r\n# \u9ed8\u8ba4\u7684\u6700\u5927\u5b57\u7b26\u4e32\u957f\u5ea6\u63a7\u5236\r\nexport OPENAI_RERANK_MAX_SIZE=1024\r\n```\r\n\r\n### \u83b7\u53d6\u6587\u672c\u5411\u91cf\r\n\r\n*\u4ee3\u7801*\r\n```python\r\nfrom openai_simple_rerank.base import get_rerank_scores\r\n\r\ns1, s2 = get_rerank_scores(\"hello\", [\"hi\", \"world\"])\r\nprint(s1, s2)\r\n```\r\n\r\n*\u8f93\u51fa*\r\n\r\n```txt\r\n0.9953891038894653 0.40583446621894836\r\n```\r\n\r\n## \u5173\u4e8e\u5b57\u7b26\u4e32\u957f\u5ea6\u63a7\u5236\r\n\r\n- \u6a21\u578b\u672c\u8eab\u4e00\u822c\u6ca1\u6709\u5b57\u7b26\u4e32\u957f\u5ea6\u63a7\u5236\u3002\r\n- \u4f46\u8fc7\u957f\u7684\u5b57\u7b26\u4e32\u4f1a\u5bfc\u5165\u6a21\u578b\u5360\u7528\u5185\u5b58\u7684\u589e\u957f\u3002\r\n- \u6240\u4ee5\u9ed8\u8ba4\u5c06\u5b57\u7b26\u4e32\u957f\u5ea6\u63a7\u5236\u5728\uff1a1024\u5b57\u3002\r\n- \u901a\u8fc7`OPENAI_RERANK_MAX_SIZE`\u8bbe\u7f6e\u9ed8\u8ba4\u6700\u5927\u5b57\u7b26\u4e32\u957f\u5ea6\u3002\r\n- \u4e5f\u53ef\u4ee5\u51fd\u6570\u8c03\u7528\u4e2d\u6307\u5b9a\u6700\u5927\u5b57\u7b26\u4e32\u957f\u5ea6\u3002\r\n- \u6ce8\u610f\uff1a\u6240\u6709\u8d85\u8fc7\u6700\u5927\u957f\u5ea6\u7684\u5b57\u7b26\u4e32\u5c06\u88ab\u622a\u65ad\u3002\r\n\r\n\r\n## \u7248\u672c\u8bb0\u5f55\r\n\r\n### v0.1.0\r\n\r\n- \u7248\u672c\u9996\u53d1\u3002\r\n\r\n",
"bugtrack_url": null,
"license": "Apache License, Version 2.0",
"summary": "\u4e00\u4e2a\u7b80\u5355\u7684rank\u6a21\u578b\u7684OPENAI\u517c\u5bb9API\u670d\u52a1\u7684\u5ba2\u6237\u7aef\u8bbf\u95ee\u5de5\u5177\u3002",
"version": "0.1.0",
"project_urls": null,
"split_keywords": [
"openai-simple-rerank"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "4f513fa523a6b24f778f59f184c388af2142bc83ef5c8b4d5c585aad9004829b",
"md5": "4c1aafd70c6d8df5822a67f966673982",
"sha256": "ff8d070921489a2b80566022f42e38983952b673962b5bb795bf83965fc5a1c7"
},
"downloads": -1,
"filename": "openai_simple_rerank-0.1.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "4c1aafd70c6d8df5822a67f966673982",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 9430,
"upload_time": "2024-12-24T14:04:51",
"upload_time_iso_8601": "2024-12-24T14:04:51.149620Z",
"url": "https://files.pythonhosted.org/packages/4f/51/3fa523a6b24f778f59f184c388af2142bc83ef5c8b4d5c585aad9004829b/openai_simple_rerank-0.1.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "85b5a53528ec4faac0faf0beca381fd85f86cdccd7050db1ea238417b52526b0",
"md5": "d9e9998c5651bcba6474337532aa7e62",
"sha256": "6d68596d8d3e7e2a257d1814f161ddfd60cdfced2b010391922981180a00bafe"
},
"downloads": -1,
"filename": "openai-simple-rerank-0.1.0.tar.gz",
"has_sig": false,
"md5_digest": "d9e9998c5651bcba6474337532aa7e62",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 8556,
"upload_time": "2024-12-24T14:04:53",
"upload_time_iso_8601": "2024-12-24T14:04:53.955475Z",
"url": "https://files.pythonhosted.org/packages/85/b5/a53528ec4faac0faf0beca381fd85f86cdccd7050db1ea238417b52526b0/openai-simple-rerank-0.1.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-12-24 14:04:53",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "openai-simple-rerank"
}
JSON