alibaba / alibaba/Logics-Parsing

建议提供官方托管 API / 百炼(DashScope)接入,减少本地部署门槛

Open
#32 1 comment 1 reaction 0 assignees View on GitHub
Dominant language
Python
Stars
1.4k
Forks
122
PR merge metrics
No merged PRs in 30d

Description

你好,感谢开源 `Logics-Parsing`。

目前 README 的主路径还是以本地权重 + `inference_v2.py` 推理为主。对于很多真实用户来说,这条路径门槛还是比较高,主要有几个问题:

1. 显存门槛较高
- 很多常见设备只有 8GB 显存,难以稳定跑通 `Logics-Parsing-v2`
- Windows 环境下依赖、加速库、推理稳定性也更容易踩坑

2. 批量文档场景更需要“服务化”而不是“单机脚本”
- 教材 / 讲义 / PPT / PDF 批量解析时,用户更希望直接调用一个稳定 API
- 对 OCR / 文档解析工具链开发者来说,官方托管 API 也更容易集成到现有系统里

3. 当前生态里,用户很自然会期待阿里云侧有对应能力
- 比如百炼 / DashScope / OpenAI-compatible API
- 哪怕第一阶段只提供基础文档解析能力,也会大幅降低使用门槛

想请教团队:
- 后续是否有计划提供官方托管 API?
- 是否考虑接入阿里云百炼 / DashScope?
- 如果已经有内部路线,是否可以给一个大致 roadmap?

如果短期内还不会提供完整托管服务,也很希望至少能有:
- 官方服务化部署方案(例如 vLLM / HTTP service)
- 标准请求 / 响应格式示例
- 面向 PDF / PPT / image 的统一 API 接入文档

我觉得这会极大提升 `Logics-Parsing` 的实际可用性和落地速度。谢谢!

Contributor guide

No contributing guide indexed for this repository

Research direction

Start with README and inference_v2.py to understand the current local-inference path and its deployment assumptions. Review the requested vLLM/HTTP service, DashScope or OpenAI-compatible API, and unified PDF/PPT/image examples as possible scopes. Done requires an agreed, implementable API or deployment scope with defined request and response behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
api, cloud
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.