Skip to main content

星火人设工具 —— 代码优化版

自我认知工具链是一个基于讯飞星火开源13B大模型做的场景训练方案,iflytekspark-13B模型已深度优化人设的数据适应,能在1000条数据上达到模型自身通用人设的覆盖及变更。

官方地址

本工具作为我的修改版主要有以下增强:

  • 代码优化,很多代码不合理的地方进行了修改
  • 性能增强,调用大模型使用了异步 io
  • 引入了 dotenv 来处理环境变量问题
  • 惯例优于配置,处理了命令行体验问题
  • 作为独立的 pip 包来发放而不是作为代码仓进行发放
  • 支持断点续传

还没有做的:

  • 目前没有兼容星火官方 websocket api,我使用了一个代理用来处理,保持和 OpenAI 的兼容性(主要是不想造轮子,后续有可能推动转换项目开源和为了更好处理异步 io
  • 本工具最终目标是给星火开放平台的 train 平台做人设工具数据生成,推荐采用开放平台的 lora 来进行设定改进

python 版本要求 3.10 以上

使用说明

  1. 安装 pip install yany
  2. 初始化项目 yany init
  3. prompt 生成 yany g p
  4. dataset 生成 yany g qa

注意,一定要修改项目目录的 .env 文件,填入自己的 key

一般来说一个项目需要进入你要的文件夹,然后执行 2、3、4 步。训练有问题请联系讯飞云平台,如果你是用开源的讯飞模型那么请自行解决。


开源协议

MIT

Metadata

Release files for yany 0.3.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for yany 0.3.0
File Size Uploaded Provenance Actions
yany-0.3.0.tar.gz 63.2 kB Without provenance Details for yany-0.3.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for yany 0.3.0
File Interpreter ABI Platform Actions
yany-0.3.0-py3-none-any.whl Python 3 none any Details for yany-0.3.0-py3-none-any.whl

Total release size: 127.5 kB

Release files / yany-0.3.0.tar.gz

Download URL yany-0.3.0.tar.gz
Size 63.2 kB
Tags Source
SHA-256 checksum
How to use checksums
8904c71263472f531f4c977e2ad51465ea783fc7c1c20bb76bed624ab3bc7262
BLAKE2b-256 checksum
How to use checksums
ecdadcb69de83a47baa786ce0ce706d81eb21c2382398328347f60ebe67be743
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/5.0.0 CPython/3.12.2

Release files / yany-0.3.0-py3-none-any.whl

Download URL yany-0.3.0-py3-none-any.whl
Size 64.4 kB
Tags Python 3
SHA-256 checksum
How to use checksums
1ef7231a68c0fb5ff31315d09960d12cf2282292fb9f3322646dcba8c24ef537
BLAKE2b-256 checksum
How to use checksums
9d93723ec979587a11fc46a4bfe578695d9dbf1c48aaa4ccef2033b071bd83a8
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/5.0.0 CPython/3.12.2

Security for yany 0.3.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page