Google Summer of Code Discussion

[tpike3]

If you are interested in supporting Mesa during the Google Summer of Code, please ask questions or other inquiries here!

GSoC Idea List

[apranav22]

Hey! Has any work been done on these ideas given in the list previously?

[Wei-cyber]

Hello. project ideas look very interesting. Should we work on the initial proposal right now or do we have to submit a contribution as mentioned in the contributor guide ?(https://github.com/projectmesa/mesa/blob/main/CONTRIBUTING.md)

[rht]

The application opens on March 18, which is ~a month from now, so you have plenty of time. Submitting small contributions is not required, but would help us in evaluating the potential contributors, given the limited number of slots.

[Wei-cyber]

Thanks

[Prithvij2004]

Hi there! I just wanted to say that I find the project ideas really interesting and appealing. I'm actually currently learning about RL algorithms myself, and I think it would be a great opportunity to contribute to the project.

[tpike3]

@Prithvij2004 Great to meet you and let us know if you have any questions

[Prithvij2004]

I have little knowledge about Reinforcement Learning, but I would like to learn more about RL. Also I am new to open source contributing. I quite didn't catch the RL problem mentioned in the tasks. It would be great if you could tell me where I could get the resources.

…

On Tue, Mar 5, 2024, 5:45 PM Tom Pike ***@***.***> wrote: @Prithvij2004 <https://github.com/Prithvij2004> Great to meet you and let us know if you have any questions — Reply to this email directly, view it on GitHub <#2046 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXB6RMAMGRH376FAEZ26KTDYWWZOJAVCNFSM6AAAAABDTYJ4HGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DMNZZGUYTQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[tpike3]

@Prithvij2004 Here are some libraries with tutorials you should find useful. Let us know if you have any more questions:

• https://openrl-docs.readthedocs.io/en/latest/
• https://keras.io/examples/rl/ ; https://keras-rl.readthedocs.io/en/latest/
• https://www.gymlibrary.dev/content/tutorials/

[Prithvij2004]

Hey, I have a doubt regarding DQN, where can I contact you? This is based on the code I have seen leading frameworks made. Can we discuss it here? Or should I mail you?

[tpike3]

Apologies just saw this you can email me at tpike3@gmu.edu or post here.

[madman024]

Hello everyone, i want to work on the network scheduler project ,to clarify i dont know a lot of things ,but i am willing to learn many things on the way, how can i get started on the project

[EwoutH]

Start looking into the current Mesa codebase, check the examples, built some models, and get a feeling about what the current limitations are around Networks and scheduling.

Also read #1912. It's an important open discussions we're currently having within Mesa.

[madman024]

yeah got it ,thanks a lot,if any need i willl reach out to you

[ankitmrj]

Greetings!

I am Ankit Srivastav , a first year student of Bachelor of Technology (Computer Science and Engineering) from MMMUT India.
I have experience in the field of MERN development. I am open source contributor and eager to contribute to esteemed organization.
I am thrilled to contribute and learn from idea Cacheable Mesa.
Please help way to contribute ,discuss and share ideas more.

Email: vaibhavsribhai@gmail.com
Ankit Srivastav

[tpike3]

Welcome @ankitmrj let us know if you have any questions

[manvendra5345]

Hi everyone!
I am Manvendra Singh Solanki from Cse Ai-Ml background and i am also interested in working in this project , I know python , java , HTML , CSS . I can adapt very fast and have great learning capacity but these thing are common what i think i have different in me is that i want to learn from thoes who are better than me and catch up to them and even beat them and always keep learning.

Email:- singhmanvendra2003@gmail.com

[tpike3]

Welcome @manvendra5345 let us know if you have any questions

[VibhuAg]

Hi all,
I'm interested in working on integrating RL libraries into Mesa. I currently work on RL research and have experience using libraries like PettingZoo for MARL. I'm looking forward to contributing to the Mesa community by bringing in my knowledge of RL and MARL!

[EwoutH]

Welcome! If you have any questions, let us know

[scramblerdoodle]

Hey!
I'm currently in my last year of studies in Applied Mathematics and, after having taken a course on Emergence in Complex Systems (in which we used ABMs to study their behaviour) during my exchange program, I became very interested in the subject!
I've been working as a Python developer for the past 6 years and I've been wanting to get into open source projects since then, and I would love to contribute to Mesa.

I'm interested in the Cacheable Mesa project idea, since I have some experience with storing states for mathematical models, and I had a few questions:
Is the main idea to follow up on the issues related to this topic, as well as continuing to implement the related forks (for example mesa-interactive and mesa-replay)? Or is it more like a catch-all of the different ideas as examples on what can be done to implement cache on Mesa?
I'd love to discuss further any points of attention related to this project idea!

As a side note, I've been already messing around in building the project locally and running the test suite, and have been looking at the issues I could start contributing with. If there are any relevant contributions I can already get started on, do let me know!

[adamamer20]

Do you have a specific template for the proposal or following the guidelines here will do?
Thanks!

[EwoutH]

There is no required template, following the guidelines is a good idea. But feel free to structure it as you see fit (form follows function).

[vmaciels]

Hi there!
I'm very interested in contributing to the scheduler project. I have a masters in network science, and although I have much to learn in the programming part, I think I can help a lot in including complex networks into the library.
About the inclusion of the scheduler as a part of a dynamic network, would the focus be on implementing it from scratch in the project itself, or using other libraries like networkx? And is the idea for it to be constrained to the spatial module, or maybe implementing a new topological module?

[EwoutH]

I think other libraries often have the preference above custom code we need to maintain.

If you can, join our dev call Tuesday. Think about use cases for such a scheduler beforehand. This feels like an idea worth discussing in person.

[EwoutH]

Going though example libraries is often a good way to get familiar to ABM use cases. However, be aware of survivor bias, things that are hard to implement might be less example for.

[Soyvor]

Hello everyone i want to contribute to the mesa Rl Project and be a part of it as i wanna be a part of a big community and also be a part of gsoc, I have experience with Reinforced Learning and i have been working with various libraries under Rl and would love to be a part of it, i just need guidance on where to start from.

[EwoutH]

Welcome! For Mesa RL, some work that has been done:

• Mesa RL #2070
• Implementing reinforcement learning into mesa[GSOC - 2024] #2075

[Jacksonxhx]

Hi, I'm Jackson Xie, currently studying computer science and math in NYU, I'm experienced in developing with python and have some knowledge about ml, nlp and application of them. I'm currently working in a Financial AI company as a software engineer. To be honest, i did not have the experience of developing in an open source project before, but that's also the reason i'm here. I'm responsible and really interested in this project, specifically the Mesa RL project. I will try this repo and play with it these days and hope to contribute to the project!

[EwoutH]

Welcome! If you have any question please let me know.

[Increshi]

Hello all, I am Rushi currently doing my Btech at the Indian Institute of Technology Tirupati. I have a good experience in Python and Cpp. Also, I have good knowledge of ML algorithms, DL, and Reinforcement learning. I have also done my internship at the Robert Bosch Center of Data Science and Artificial Intelligence, IIT Madras, where I worked on Gen AI. I am also familiar with Python Libraries Numpy, Matplotlib, pandas, and ML frameworks like Pytorch, Tensorflow, and Keras.I also have experience with Git and Github. I am very interested in the MesaRL Project as it explores RL which I like most.And I love playing with RL Algos like PPO, Q learning etc. Looking forward to working with you.

[EwoutH]

Welcome! If you have any questions please let us know.

[partizanos]

Hello, I am Dimitrios Proios, currently a 3rd-year PhD student in the Faculty of Medicine at the University of Geneva in Switzerland, with a focus on representation learning using heterogeneous temporal graph neural network models. I found the Network Scheduler proposal fascinating and aligned with my research interests. I am proficient in Python and primarily use PyTorch for machine learning, though I have also worked with TensorFlow. Additionally, I have utilized JAX to teach reinforcement learning (Q-learning, bandit algorithms) at the Mediterranean Machine Learning Summer School in 2024. My experience includes working at CERN and in financial institutions as a web developer, data engineer, and quantitative developer with 8 years of work experience in both research and industry (LinkedIn, Google Scholar), it would be a pleasure to contribute to the project in the coming months through source contribution, as well as provide maintenance and support as an active member of the open-source community.

[EwoutH]

Welcome! If you have any questions please let us know.

[madman024]

hey @tpike3 @EwoutH @rht sry to tag u all ,but i was filling my gsoc application but now theres no submission form,ig theres still time as its 4:47 utc,what should i do.

[EwoutH]

Can you post a screenshot of what you do see? Are you properly logged in? If correct, you should have another 56 minutes.

[EwoutH]

Maybe try to see if you can reach GSoC support in some way.

[madman024]

@EwoutH hey i wasnt able to fill ,i filled the proposal form and then it reloded and the button and form to submit proposal were not there i dont know what was the issues .thanks for your msg.btw could u just go through my proposal and give me feedback,that would be helpful for me.