Updating documentation of "examples/robot-cityscapes-synthia/lifelong_learning_bench/semantic-segmentation/README.md"

[AryanNanda17]

I am an open-source contributor who is interested in contributing to the project Multimodal Large Model Joint Learning Algorithm: Reproduction Based on KubeEdge-Ianvs. Issue #123
What would you like to be documented:
In this Readme, there are a few issues.

• instead of using python to run different commands, it should be python3. For example instead of python -m pip install ./examples/resources/third_party/*, it should be python3 -m pip install ./examples/resources/third_party/*
• Also, before running this command, python -m pip install --upgrade pip, pip version lower than 24.* should be installed. (with python3.8.*)

Here no dataset download command is mentioned and a direct unzip command is run

• Also, there are few pre-requisites to download before running these steps like installing git, wget, unzip commands using apt-get

Why is this needed:

I volunteer to make these changes. This will improve the documentation. I can do a PR after the maintainers approval.

Also, please give the right command to download the mdil-ss dataset.

[AryanNanda17]

Also, @CreativityH @hsj576
Please suggest if there is any more improvements in documentation with which I could help with. I can try to run different files that you mention, it will be a good start and I can understand better along the way.
Thanks

[CreativityH]

@hsj576 @MooreZheng These issue and enhanced PR make sense. Please consider the improved Readme. Also, please provide the right link or command to download the mdil-ss dataset.

[MooreZheng]

Message received. Newcomers are very welcome.

Agree that a website link could be added in the readme.

1. @AryanNanda17 might want to help solve the issue by contacting @qxygxt @Frank-lilinjie @hsj576 . This is an advanced research item from past events. It might not be so user-friendly to all newcomers and the authors might not respond very soon.
2. Besides, it seems that this dataset is a collection of public datasets of cityscape and synthia. So one can also fix it by exploring the two datasets.

Newcomers might try to start the journey with https://ianvs.readthedocs.io/en/latest/guides/quick-start.html and then try fixing the "good first issues" of ianvs.

[AryanNanda17]

Hello @MooreZheng,
Thanks for your response.
I followed the instructions mentioned in the Quick Start guide. I faced the same issues as @octonawish-akcodes (a bunch of compatibility issues, it says to use python3.6.9 but when this is used, other packages give an error saying that they need a higher version of Python) and so on. The quick-start guides need an update.

[AryanNanda17]

Should I share the errors here so I can update the guide with your help and the community's input?

[FuryMartin]

Hello @MooreZheng, Thanks for your response. I followed the instructions mentioned in the Quick Start guide. I faced the same issues as @octonawish-akcodes (a bunch of compatibility issues, it says to use python3.6.9 but when this is used, other packages give an error saying that they need a higher version of Python) and so on. The quick-start guides need an update.

Hi, I've also discovered that Ianvs currently has some issues with incomplete dependencies, and I am trying to fix it in #132.

If you want to run an Ianvs project, I suggest referring to #106, where I wrote the running guide for examples/robot/lifelong_learning_bench/semantic-segmentation. Hopefully this helps you a bit.