OpenMMLab CodeCamp 2023, Apply NOW! #1719
Description
Introduction
超级视客营再度启航，展现编程实力，共塑开源未来！
第二期超级视客营提供 10 个课题方向、150+ 不同难度课题任务，包括但不限于基础架构、预训练、目标检测、AIGC 等多个领域。活动算力由北京超级云计算中心提供支持，为大家的开发保驾护航。
The OpenMMLab CodeCamp has BEGUN! Harness programming skills and shaping the future of OpenMMLab.
Oriented to developers worldwide, released 150+ tasks in 10 popular areas, including fundamental framework, pre-training, AIGC and more.
Process
-
进入超级视客营活动页面 >> 查看任务详情 >> 选择感兴趣的方向（如MMEngine）>> 选择任务难度 >> 选择意向任务（可填报 3 个志愿）
-
点击立即报名
-
提交报名申请表
-
等待结果通知：站内信、邮箱、短信
- 申请成功 >> 开始任务，在成功通知的结果中将获取 TaskID
- 申请失败 >> 再次提交任务申请
-
提交PR，PR title 规范如下： '[CodeCamp2023-TaskID] xxx'（例如：TaskID为 325 的任务 Find the proper learning rate，PR 的 Title 应为 '[CodeCamp2023-325] Find the proper learning rate'）
-
PR经过review，合入代码库，该任务标记完成
-
即可领取下一个任务
-
Enter the Event Page >> View task details >> Choose the direction (such as MMEngine) >> Choose task difficulty >> Choose the desired task (you can have 3 choices)
-
Click the Register Now
-
Wait for the result notification: on-site messages, email
-
Working on the task
-
Submit PR, PR title format: '[CodeCamp2023-TaskID] xxx' (For example: For TaskID 325 "Find the proper learning rate", the PR Title should be '[CodeCamp2023-325] Find the proper learning rate')
-
Once the PR has been reviewed and merged into the codebase, the task will be marked as complete
-
You can then pick up the next task
MMPreTrain Task List
| 任务 | Task | Difficulty |
|---|---|---|
| 支持 ImageNet-A,R,S,CSupport ImageNet-A, R, S, C in MMPreTrain | Add support for the ImageNet Out-of-Distribution (OOD) dataset in MMPreTrain. | Medium |
| 新版 config 适配 ResNet | The new version of config adapts ResNet algorithm: |
Easy |
| 新版 config 适配 BeiTv2 | New Version of config Adapting BeitV2 Algorithm: |
Easy |
| 新版 config 适配 MAE | New Version of config Adapting MAE Algorithm: |
Easy |
| 新版 config 适配 ConvNeXt | New Version of config Adapting ConvNeXt Algorithm: |
Easy |
| 新版 config 适配 Swin Transformer | New Version of config Adapting Swin Transformer Algorithm: |
Easy |
| 新版 config 适配 ViT | New Version of config Adapting Vision Transformer Algorithm: |
Easy |
| 新版 config 适配 MobileNet | New Version of config Adapting MobileNet Algorithm: |
Easy |
| 多模态数据集文档补充 - COCO Retrieval | Support retrieval tasks with COCO Retrieval dataset | Easy |
| 多模态数据集文档补充 - COCO Caption | Support image captioning task with COCO Caption dataset | Easy |
| 多模态数据集文档补充 - RefCOCO | Support detection task with RefCOCO dataset | Easy |
| 多模态数据集文档补充 - ScienceQA | Support question answering task with ScienceQA dataset | Easy |
| 多模态数据集文档补充 - COCO VQA | Support question answering task with COCO VQA dataset | Easy |
| 支持 mPLUG-owl 多模态算法推理 | Support for mPLUG-owl multimodal algorithm inference | Hard |
You will gain
- 核心开发者 1 对 1 指导 One-on-one tutoring from core developers of the codebase
- 精美电子奖品 Electronic prizes
- 结营证书及周边 OpenMMLab certificate and swag
- 面试直通车 Fast pass of interview
Due to delivery issues,the majority of physical prizes are limited to mainland of China.
Group discussion
WeChat：
Discord：https://discord.gg/KuWMWVbCcD
More Details：https://openmmlab.com/activity/codecamp