Calling for volunteers for developing cool features! 🚀

[HAOCHENYE]

Introduction

Interested in deeply participating in OpenMMLab projects? Want to learn more about OpenMMLab's awesome tools without wasting plenty of time reading docs? The First OpenMMLab Codecamp has begun!! We provide more than a hundred tasks from seventeen research areas for you to choose from. Whether you are a novice in AI or a senior developer, there are suitable tasks for you to participate in. We will provide quick responses and full guidance to help you smoothly complete those tasks and upgrade to a core contributor of OpenMMLab. We cooperate with Beijing Super Cloud Center to provide computing power support.

How to participate?

Select the task you are interested in and submit registration here. We will inform you in three days if you have enrolled for the tasks, and then you can formulate the task plan with tutor and start development ! Once your PR has passed preliminary review, you can apply for the next task or just wait for the award!
More details: OpenMMLab Activity page。

Task description

MMEngine, as the new generation training framework of OpenMMLab, will continue to absorb and integrate attractive cutting-edge technologies, however, our own power is limited, we hope to build an active community and leverage the power of the community to make MMEngine better and better. With the help of this event, we want to actively interact with the community to develop some very interesting features. We have graded some of these tasks as follows:

Task	Discription	Related skills	Difficulty	Credits	Status
Support Segmind backend	Implement Sigmindbackend based on BaseVisBackend. References: SegmindLoggerHook step1: Implement SegmindBackend step2: Update unit test step3: Update the visualization result in PR message	Python, PyTorch	⭐️⭐️⭐️	10	WIP
Support MLFlow backend	Implement MLFlowBackend based on BaseVisBackend References: MlflowLoggerHook step1: Implement MLFlowBackend step2: Update unit test step3: Update the visualization result in PR message	Python, PyTorch	⭐️⭐️⭐️	10	WIP
Support Dvclive backend	Implement DvcliveBackend based on BaseVisBackend References: DvcliveLoggerHook step1: Implement DvcliveBackend step2: Update unit test step3: Update the visualization result in PR message	Python, PyTorch	⭐️⭐️⭐️	10
Support Neptune backend	Implement NeptuneBackend based on BaseVisBackend References: NeptuneBackend step1: Implement NeptuneBackend step2: Update unit test step3: Update the visualization result in PR message	Python, PyTorch	⭐️⭐️⭐️		WIP
Support ClearML backend	Implement ClearMLBackend based on BaseVisBackend References: ClearMLLoggerHook step1: Implement ClearMLBackend step2: Update unit test step3: Update the visualization result in PR message	Python, PyTorch	⭐️⭐️⭐️	10	WIP
Support ProfilerHook	Implement ProfilerHook based on Hook References: ProfilerHook step1: Implement ProfilerHook step2: Update unit test step3: Update the effect in PR message	Python, PyTorch	⭐️⭐️⭐️	10	WIP
Support early stop in MMEngine	Support stop training when validation metric reaches a threshold. step1: Design and implement early stop step2: Update unit test	Python, PyTorch	⭐️⭐️⭐️	10	WIP
Support ApexOptimWrapper	Implement ApexOptimWrapper based on OptimWrapper step1: Implement ApexOptimWrapper step2: Update unit test	Python, PyTorch	⭐️⭐️⭐️⭐️	30	WIP
Support ReduceLROnPlateau	Implement ReduceLROnPlateau based on ParamScheduler References: torch.optim.ReduceLROnPlateau step1: Implement ReduceLROnPlateau step2: Update unit test step3: Update the visualization of learning rate in PR message	Python, PyTorch	⭐️⭐️⭐️⭐️	30	WIP

Sign Up Here : application form 🚀 🚀 🚀

By the way, we strongly encourage you to publish your experience on social media like medium or twitter with tag "OpenMMLab Codecamp" to share your experience with more developers! 😆 😆 😆

Discussion group: discord link

Welcome to join the discussion below or in discord. Come to take the challenge, and become a contributor to the OpenMMLab !
🥳 🥳

[fcakyon]

Hello @HAOCHENYE, @zhouzaida, @hhaAndroid. Thanks for the great effort on all Open MMLab packages! We need your time to discuss some details.

@Blaizzy is from Neptune.ai and wants to discuss the roadmap of MMDetection and MMEngine with you. What questions do you have in mind @Blaizzy?

[Blaizzy]

Thank you for the warm introduction @fcakyon!

Nice to e-meet you @HAOCHENYE, @zhouzaida, and @hhaAndroid.

[Blaizzy]

Do you mind sharing when will MMDetection 3.x be released? And what are the main breaking changes from MMDetection 2.x?

[Blaizzy]

Additionally, I would like to introduce my colleague @kshitij12345 who worked on the Neptune integration with the current MMDetection 2.x to the chat. He will be also working on the new integration with MMEngine 🚀

[kshitij12345]

Thanks for the intro @Blaizzy @fcakyon!

Hello @HAOCHENYE, @zhouzaida, @hhaAndroid,

First of all, thank you for maintaining the great packages under the open-mmlab umbrella.
It would be great to discuss the road-map for packages under open-mmlab so that Neptune can seamlessly work with them. To that end, I was wondering if we should discuss this in a separate issue to keep this issue clean (or somewhere else?)

Thank you very much!

[HAOCHENYE]

Welcome @kshitij12345 @Blaizzy @fcakyon 🎅🎅! Also, appreciate the fancy experiment track supported by Neptune 😝, and we're very willing to collaborate with you to promote Neptune and OpenMMLab series repositories. I've opened a new issue for discussing the roadmap, and we may shift there🚀

[ccomkhj]

For those waiting for the official MlflowLoggerHook, this is the temporary version I'm currently using.
https://github.com/ccomkhj/MlflowLoggerhook/