Issue #283 - The ML Engineer ??
Alejandro Saucedo
Tech Executive @ Zalando | Chair/Advisor @ UN, ACM, LF, etc | Join 60k+ ML Newsletter
Thank you for being part of over 60,000+ ML professionals and enthusiasts who receive weekly articles & tutorials on Machine Learning & MLOps ?? You can join the newsletter for free at https://ethical.institute/mle.html ?
If you like the content please support the newsletter by sharing with your friends via ?? Email, ?? Twitter, ?? Linkedin and ?? Facebook!
This week in Machine Learning:?
The many ways to deploy a machine learning model from the team at Outerbounds: A great overview for production ML practitioners on the diverse deployment strategies for ML/AI models, focusing on the critical considerations of scale, reliability, and iteration speed. It is often key to consider the broad range of technical requirements based on the application needs - i.e. batch vs real time, reusable vs specialized, modalities, etc. Great practical examples and conceptual frameworks to identify the most suitable deployment approach.
xLSTM introducing the Extended Long Short-Term Memory as a challenger for the ever growing wave of transformer architectures: European innovation introduces an advanced version of the traditional LSTM model by incorporating exponential gating and innovative memory structures to address its limitations. These modifications allow xLSTMs to perform competitively with contemporary Transformer models in language processing tasks, showcasing improved capability in handling complex memory operations and scaling to large model architectures.
Llama 3 implemented in pure NumPy - what better way to learn a concept than by implementing it: Great practical deep dive implementing Llama 3 model using only NumPy, demystifying the underlying model's architecture and nuances. Some of these include key components such as RoPE positional encoding, RMSNorm, and Scaled Dot-Product Attention, alongside optimizations like KV Cache for efficiency. This implementation serves as a great learning resource for machine learning practitioners interested in understanding the details under the hood.
META AI Research presents a scaling law for large-scale Recommendation Systems: As recsys grow in adoption and presence across tech companies, the need grows for a scaling law similar to those observed in large language models to understand relationships between considerations such as resources and limits. This design enables the model to effectively and efficiently scale by capturing any-order feature interactions through progressively deeper and wider layers.
GPUs Go Brrr - or how to optimize AI compute on GPUs: Practical strategies and philosophical shifts are necessary to maximize hardware utilization in the world of machine learning compute. Some key best practices include leveraging asynchronous matrix multiplication instructions in GPUs directly from shared memory, and managing the quirks of shared memory to minimize latency and bank conflicts. This article provides an intuitive deep dive into how to streamline complex CUDA kernel programming, making it more accessible and efficient for developers working with intricate ML algorithms.
领英推荐
Upcoming MLOps Events
The MLOps ecosystem continues to grow at break-neck speeds, making it ever harder for us as practitioners to stay up to date with relevant developments. A fantsatic way to keep on-top of relevant resources is through the great community and events that the MLOps and Production ML ecosystem offers. This is the reason why we have started curating a list of upcoming events in the space, which are outlined below.
Upcoming conferences where we're speaking:
Other upcoming MLOps conferences in 2024: ?
In case you missed our talks:
Open Source MLOps Tools
Check out the fast-growing ecosystem of production ML tools & frameworks at the github repository which has reached over 10,000 ? github stars. We are currently looking for more libraries to add - if you know of any that are not listed, please let us know or feel free to add a PR. Four featured libraries in the GPU acceleration space are outlined below.
? If you know of any open source and open community events that are not listed do give us a heads up so we can add them!
OSS: Policy & Guidelines
As AI systems become more prevalent in society, we face bigger and tougher societal challenges. We have seen a large number of resources that aim to takle these challenges in the form of AI Guidelines, Principles, Ethics Frameworks, etc, however there are so many resources it is hard to navigate. Because of this we started an Open Source initiative that aims to map the ecosystem to make it simpler to navigate. You can find multiple principles in the repo - some examples include the following:
If you know of any guidelines that are not in the "Awesome AI Guidelines" list, please do give us a heads up or feel free to add a pull request!
About us
The Institute for Ethical AI & Machine Learning is a European research centre that carries out world-class research into responsible machine learning.
? Infrastructure Engineer ? DevOps ? SRE ? MLOps ? AIOps ? Helping companies scale their platforms to an enterprise grade level
4 个月Indeed, the Machine Learning ecosystem is abuzz with innovation! Have you explored any of these topics in-depth? So much to learn and apply. Alejandro Saucedo
AI Executive, GenAItechLab.com
4 个月Thank you for sharing! See also how to build secure, local LLMs with Ollama, at https://mltblog.com/3wxFfoS