aszerW 59c03516e4 feat(router): 集成NVIDIA多头分类器实现3-tier智能路由 ...

- 新增nvidia_router.py: 手动加载NVIDIA prompt-task-and-complexity-classifier模型
- DeBERTa-v3-base backbone + 8个分类头(task_type/creativity/reasoning/domain等)
- 综合多维度评分实现simple/medium/complex三级路由
- 映射: simple->qwen-flash, medium->qwen-plus, complex->qwen-max
- main.py切换到NVIDIA路由替代RouteLLM BERT二分类
- 移除LiteLLM依赖解决版本冲突,使用原生httpx调用
- 版本升级至v0.3.0

2026-04-18 01:21:31 +08:00

docs

docs: add LLM router open source research report

2026-04-17 23:59:18 +08:00

.env.example

feat: add Qwen (DashScope) support as default provider

2026-04-17 23:47:06 +08:00

.gitignore

feat: implement MVP LLM router service

2026-04-17 23:33:43 +08:00

bert_router.py

feat: integrate RouteLLM BERT router for intelligent query classification

2026-04-18 00:12:51 +08:00

config.py

feat: add Qwen (DashScope) support as default provider

2026-04-17 23:47:06 +08:00

main.py

feat(router): 集成NVIDIA多头分类器实现3-tier智能路由

2026-04-18 01:21:31 +08:00

nvidia_router.py

feat(router): 集成NVIDIA多头分类器实现3-tier智能路由

2026-04-18 01:21:31 +08:00

requirements.txt

feat(router): 集成NVIDIA多头分类器实现3-tier智能路由

2026-04-18 01:21:31 +08:00

X402生态竞品技术架构深度调研-学术映射与生产实现对比.md

docs: add X402 ecosystem technical architecture research report

2026-04-17 23:24:55 +08:00

Description

No description provided

139 KiB

Languages

Python 92.2%

Dockerfile 4.6%

Shell 3.2%