Skip to content

[黑客松10th·文心伙伴] 周报 #24 nianliunian 2026.05.14 #608

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
nianliunian wants to merge 1 commit into
base: main
Choose a base branch
from
+41 −0
Open

[黑客松10th·文心伙伴] 周报 #24 nianliunian 2026.05.14 #608

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
41 changes: 41 additions & 0 deletions ...ERNIEPartner/ERNIEPartner_24_nianliunian/[WeeklyReport]2026.04.29~2026.05.10.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,41 @@
### 认领者 GitHub ID
nianliunian

### 赛题信息
- **进阶任务序号**：#24
- **赛题名称**：PaddleOCR-VL-1.5 在此芯 P1 芯片上的端侧部署与优化
- **关联厂商**：此芯 & Arm

### 本周工作

1. **环境搭建**
- 在此芯上搭建llama.cpp的编译环境：编译cpu和gpu，对比推理性能
- 通过性能日志，排查llama推理时的性能瓶颈
- 在本地搭建模型转换和开发测试环境：将模型从 safetensors 转换为 onnx进行推理
- 编写推理代码

2. **模型调研**
- 下载相关模型：paddleocr-vl-1.5

3. **问题与解决**
- 问题：llama运行模型较慢，主要瓶颈在图像embedding时
- 解决：考虑从原始模型中拆解出vision部分和llm部分，使用onnx边缘部署的优势，分别对visionh部分和llm部分完成推理



### 下周计划

1. 调试、完善推理代码，对给定的图像进行测试验证

### 当前阻塞（无则填"无"）

- 无

### 交付物进展

| 交付物 | 状态 | 备注 |
|--------|:----:|------|
| RFC 文档 | ✅ 已完成 | - |
| 代码实现 | 🔄进行中| - |
| README | 🔄进行中| - |
| 演示视频/截图 | ⬜未开始 | - |