integrated fine-tuning platform for lightweight vlmOCR models
Project description
Kalorda
轻量vlmOCR模型一站式微调平台
Kalorda是一个轻量vlmOCR模型微调集成平台,前端采用Typescript+Vue3+Vite,后端采用Python+FastAPI+ms-swift+vLLM构建,提供针对主流轻量vlmOCR模型的数据二次标注、微调训练、对比测试等一站式综合解决方案。
🚩安装使用
1、新建虚拟环境
# 使用 conda 新建虚拟环境
conda create -n kalorda python=3.12 -y
# 激活(切换)虚拟环境
conda activate kalorda
2、安装命令
pip install kalorda
# 或指定阿里云镜像源进行安装
pip install kalorda -i https://mirrors.aliyun.com/pypi/simple/
3、启动命令
kalorda --port 8800
可选启动参数:
--host:指定主机地址,默认值为0.0.0.0--port:指定端口号,默认值为8800--gpu-devices:指定允许使用的GPU设备索引(从0开始),默认值为空表示不限制(即全部GPU都可使用),多个GPU索引用逗号分隔,例如--gpu-devices 0,1,2--workers:指定工作进程数(至少要2个工作进程),默认值为2--log-level:指定日志级别,默认值为info
系统和硬件条件:
- Linux操作系统(Windows下请安装wsl2 ubuntu子系统)
- Python虚拟环境管理工具(推荐使用miniconda3或uv)
- 至少一张Nvidia GPU显卡,显存16G或以上,已安装显卡驱动及CUDA(非Nvidia显卡当前暂不支持,等后续)
- 硬盘空间:50GB或以上
💡联系交流
GitHub/Issues:https://github.com/vlmOCR/Kalorda/issues
微信:
(扫码添加微信,备注:kalorda,邀您加入群聊)
📜License
Kalorda项目基于Apache-2.0协议开源,您可以在遵守协议的前提下自由使用、修改和分发本项目。
Copyright (c) 2025-present, Kalorda
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file kalorda-0.1.1.tar.gz.
File metadata
- Download URL: kalorda-0.1.1.tar.gz
- Upload date:
- Size: 24.3 MB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
1beca9b4eda34dbe70845472a4eb5c2e3d2e8862e0c8f4af0b19ea8eb66e08c1
|
|
| MD5 |
ff8ce0cbf859ccb465900540f0207a14
|
|
| BLAKE2b-256 |
f52f3b34b863debfb7d68be1d64668f761490cf0645b857c1934e14824739df4
|
File details
Details for the file kalorda-0.1.1-py3-none-any.whl.
File metadata
- Download URL: kalorda-0.1.1-py3-none-any.whl
- Upload date:
- Size: 24.4 MB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
cbf01b6ed2e620944a4a932aa414275c9a969a72dfad4c4163013c8dafce6a70
|
|
| MD5 |
b72a9a87b86ee3bb7771c51209bc7043
|
|
| BLAKE2b-256 |
d80cd742a3983886a31e55be93971d32a943635702ab1f68a47344610aff6ad7
|