Introduce yourself! #35
Replies: 11 comments 6 replies
-
|
Hello, my name is Jeremy Ellis. You can find me on GitHub at https://github.com/hpssjellis and on LinkedIn at https://www.linkedin.com/in/jeremy-ellis-4237a9bb/ I've been creating neural networks for nearly 30 years and teaching high school machine learning for 8 years. Where I lack in book writing skills, I excel in simplifying ML concepts for teaching. In the cs249r_book, several chapters have discussed the security advantages of tinyML. You can explore these insights in the following links: embedded_sys.qmd Line 61 in d9c33af embedded_ml.qmd Lines 154 to 156 in d9c33af and also Line 204 in d9c33af However, it's worth noting that these chapters don't seem to mention that most tinyML models are created on cloud platforms. This inherently poses a security risk, as data is transferred during the training process, and there are questions about the handling of that data while it resides on the cloud. My particular area of interest revolves around streamlining the process of efficiently developing client-side machine learning models for tinyML. The question at hand is whether the construction of tinyML models is as secure as we assume, and if there are alternative approaches worth considering? |
Beta Was this translation helpful? Give feedback.
-
|
Hey @hpssjellis! Sorry that this message missed my inbox, I have been overwhelmed. Thanks for jotting down the points. I agree there is room for improvement. As we are working through the chapters, these points are covered in the later chapters, or at least I'd like to think so. Please continue to look through the other chapters, as they become available. Feel free to make suggestions or edits, as you see fit and issue a PR if you'd like to. We are open for contributions! Some of the specific things about security etc. will be covered in the on-device learning and security/privacy chapters 👍 |
Beta Was this translation helpful? Give feedback.
-
|
Hi all, Jared Ping here from South Africa. Group Technical Manager at Digital Matter, where we have a huge focus on the energy efficiency of battery-powered tracking devices in a range of applications (TinyML included!). Currently pursuing a part-time MSc(Eng) Electrical Engineering at the University of the Witwatersrand, specialising in operational optimisation of TinyML systems. Very excited to see the TinyML community grow, and this book is a great resource for introducing peers to it! |
Beta Was this translation helpful? Give feedback.
-
|
Hello everyone, My name is Ethan Wang, a Master's student in Software Engineering at North China Electric Power University. My journey in the tech world is driven by a profound passion for operating systems (OS) and the burgeoning field of Tiny Machine Learning (TinyML). The challenge of maximizing the capabilities of limited resources to achieve seemingly limitless tasks fascinates me. Currently, I am interning at Megatronix, where I am honing my skills in system stability, a crucial aspect of software engineering that aligns closely with my interests in OS and TinyML. My Main Interests:
Future Aspirations:My immediate goal is to deepen my understanding and expertise in OS and TinyML through practical experiences and academic research. However, looking ahead, I am also setting my sights on opportunities to study and work abroad. I believe that pursuing further education and professional experiences abroad will not only broaden my horizons but also provide me with a unique perspective on global technological challenges and innovations. Contributions to the Community:
Additional Links:
I am looking forward to engaging with this community, learning from you all, and contributing where I can. Thank you for welcoming me into this vibrant space! |
Beta Was this translation helpful? Give feedback.
-
|
Nice to see that this is an active repo and work in progress with commits as latest as yesterday! |
Beta Was this translation helpful? Give feedback.
-
|
Welcome Seshu, please feel free to share any thoughts and feedback you may
have. We have some exciting new updates coming through!
Vijay Janapa Reddi, Ph. D. |
John L. Loeb Associate Professor of Engineering and Applied Sciences |
John A. Paulson School of Engineering and Applied Sciences |
Science and Engineering Complex (SEC) | 150 Western Ave, Room #5.305 |
Boston, MA 02134 |
Harvard University | Email ***@***.***> | Website
<http://scholar.harvard.edu/vijay-janapa-reddi> | Google Scholar
<https://scholar.google.com/citations?hl=en&user=gy4UVGcAAAAJ&view_op=list_works&sortby=pubdate>
| Edge Computing Lab <https://edge.seas.harvard.edu> | Schedule a Meeting
<https://scholar.harvard.edu/vijay-janapa-reddi/schedule> | Admin
<https://scholar.harvard.edu/vijay-janapa-reddi/contact> |
…On Fri, Mar 22, 2024 at 8:17 PM, Seshu-R ***@***.***> wrote:
Nice to this that this is an active repo and work in progress with commits
as latest as yesterday!
Just started reading.
Hope to see Exercises and Labs soon. Thanks in advance for all the efforts
and helping the broader community.
—
Reply to this email directly, view it on GitHub
<#35 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABT6DFG74G3YRGUCOFU5RKLYZTCZ3AVCNFSM6AAAAAA6F4456CVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DQOBTGYZDO>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Hello everyone, You can find me on:- |
Beta Was this translation helpful? Give feedback.
-
|
If you're interested in robotics, I strongly encourage you to engage in practical deployment scenarios of AI. Robotics and AI require embedded machine learning technologies and capabilities. Both these domains are computationally very intensive and they're also severely power constrained, therefore HAI technologies are going to be quite critical for them. |
Beta Was this translation helpful? Give feedback.
-
|
Hello Professor Reddi and everyone! 👋. I'm Hardik Agarwal, an undergraduate student at the Indian Institute of Technology, Delhi. My interest in ML ignited when I saw how effective it can be in tackling real-world problems with surprising efficiency and precision. Over time, I began to explore many aspects of machine learning, and I am presently most interested in TinyML. The idea of deploying machine learning models on resource-constrained devices and optimizing them for edge applications is both daunting and interesting to me. In the future, I hope to contribute by developing more efficient AI models and robotic systems that can operate seamlessly in real time and with minimal energy consumption. I’m really passionate about the intersection of machine learning, robotics, and optimization, and I see this as the key to unlocking new possibilities in various industries. I've been reading your Machine Learning Systems book, and I must say that it's quite informative and useful. Your focus on the practical aspects of ML deployment and system optimization, which are frequently disregarded in other resources, fills a critical need. Thank you for creating such a useful resource! I'm looking forward to networking with everyone and learning more about ML systems. I’m eager to learn from your expertise and insights as well as from my fellow students who share my passion for machine learning and robotics. I will be happy to connect here: |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for reaching out! I'm always excited to see students engaging with this resource. You are the reason I feel inspired to keep developing it. I remember being deeply motivated to study computer architecture as a student because of Hennessy and Patterson's textbook on computer organization and advanced computer architecture. If you are interested in TinyML -- checkout tinyML on HarvardX. There are many other resources I can recommend, but this is a great starting point! You can take it for free and if you need a kit for Course 3 let me know. |
Beta Was this translation helpful? Give feedback.
-
|
Hello Everyone, 👋 I'm Poojak Patel, a student at Illinois Mathematics and Science Academy. Since a young age, I have wondered how cars operate, which has grown into a broader interest in how machine learning could be used to enhance autonomous systems. Currently, I am interested in how we can create autonomous systems using machine learning that can help robots learn in real time. The idea that machines can both react to what's happening around them while learning from their experiences just like how a person would is fascinating. Also, after reading more about TinyML in Professor Reddi’s Machine Learning Systems book, I have been intrigued about its potential as it allows machine learning models to run on small devices with low power. I think, just like machine learning, TinyML can lead to more innovative solutions in the future. I am excited about the possibilities that this kind of technology could bring and how it would be able to handle current existing challenges, making a difference in people's lives. I would love to connect with others who are also interested in machine learning, robotics, or any related topics. Feel free to reach out to me by email ([email protected]) and see my past projects on GitHub (github.com/PatelPoojak). Looking forward to engaging with all of you in this community! Best Regards, |
Beta Was this translation helpful? Give feedback.
-
|
Hi Poojak 👋 Nice interests. Embodied AI is expected to be a significant topic in the coming years. I believe the next major challenge will involve how language models, like ChatGPT, can begin interacting with physical sensors in the real world. This interaction will provide them with a wealth of new information and data to process. Therefore, I genuinely think we are only at the tip of the iceberg. |
Beta Was this translation helpful? Give feedback.
-
|
Hi everyone! I’m Mimi, and I currently work in helpdesk but am aiming to transition into software development. I'm starting with front-end development through a Meta course and am really enjoying learning the basics. I'm eager to expand my skills and learn from this community, so any tips or advice would be greatly appreciated as I make this career shift! I would love to connect! |
Beta Was this translation helpful? Give feedback.
-
|
Hello and welcome! A general word of advice -- it's important to complement your knowledge with AI tools, as they will play a major role as a copilot in assisting us with programming tasks. Therefore, becoming proficient with these tools while still learning the fundamentals (which is equally important) will be essential for all of us moving forward. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for the warm welcome and the advice! I completely agree that AI tools are becoming an essential part of programming. I'll definitely look into learning more about them while continuing to strengthen my fundamentals. It’s exciting to see how they can complement our skills and enhance our efficiency! Any suggestions or resources I can reference to start learning about AI tools? |
Beta Was this translation helpful? Give feedback.
-
|
Hi all! My name's Ryan. I'm currently working at Apple doing cloud/backend engineering with a focus now on testing and system integration. No practical ML experience at this time. Working at Apple has inspired a data-privacy-centric view in me. That, combined with an increasing personal desire to spend less time on socially connected, networked devices has spurred an idea in me which involves using federated machine learning on a dis-connected personal device. My desire is to create a simple AI assistant that is vocally interacted with (no screen) and dis-connected from the internet (on-device learning with data privacy). Soon after I had this thought, I came across this book and thought it would be a great way to dive into ML, especially after hearing about TinyML and its relevance to the on-device learning I want to do. I've never worked with hardware, so this will be a fun and new challenge. Fun fact: Another passion project of mine is electronic music production. |
Beta Was this translation helpful? Give feedback.
-
|
Hi everyone, I’m Danjie, an undergraduate at University of Toronto. My first exposure to ML systems came when I was trying to integrate Apple’s Neural Engine with a personal project. I managed to get BERT running on my M1 laptop (with some reshaping and some other steps that I don’t really remember). The results were incredible, the Neural Engine achieved seven times the throughput of the M1’s GPU while consuming just 5 watts of power! On top of that, I have a close friend working at Annapurna Labs in Toronto on AWS’s Inferentia service. He told me they were working on getting Llama3 405b to run on Inferentia chips. It’s exciting to see how ASICs are rapidly gaining traction in the industry and generating millions in revenue for AWS. I’m particularly passionate about the co-design of hardware and software for AI acceleration. My background includes experience in computer architecture, assembly language, and high-level machine learning with PyTorch. This background allows me to see that there’s immense potential to improve efficiency across all kinds of applications, though I’ve often struggled to find resources on this topic. Thanks so much for the course! All the best, |
Beta Was this translation helpful? Give feedback.
-
Hi everyone! 👋
I’m Prof. Vijay Janapa Reddi, the author and editor of this resource, and I’m passionate about making foundational technologies like ML/AI accessible to everyone. Along with my students and TAs, I created this resource because we noticed a significant gap: while there are countless books and materials on deep learning theory and concepts, there’s a shortage of resources focused on the ML systems side.
If ML algorithm developers are like astronauts, ML systems engineers are the rocket scientists and mission control specialists who get them to space and keep the mission on track. 🚀 📡
The critical aspects of ML—optimizing models for specific hardware, deploying at scale, and ensuring efficiency and reliability—are often overlooked. These elements transform ML models into real-world applications. Yet, many practitioners struggle due to a lack of available knowledge, not interest.
That’s a little bit about me and my passion.
Now, I’d love to get to know you—our readers and fellow enthusiasts!
Please introduce yourself below and share:
Whether you’re just getting started or have years of experience—your perspective is important to us. Let’s connect, share our insights, and build a vibrant community around machine learning systems together. 🚀
Sincerely,
vj.
Beta Was this translation helpful? Give feedback.
All reactions