Issue #305 - The ML Engineer ??
Alejandro Saucedo
Tech Executive @ Zalando | Chair/Advisor @ UN, EU, ACM, 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:
Almost 60% of real-time machine learning is still powered by FastAPI/Flask or Custom Wrappers - it seems the MLOps ecosystem is not yet consolidated, showing huge opportunity in this space: These are really important insights from 2024 Survey on The State of Production ML; we have designed the questions to provide meaningful insights on the current landscape of production ML in 2024 - if you have a chance we would be grateful if you could spend a few minutes on the survey, as you'll contribute valuable information about the machine learning tools and platforms you use in your production ML development. Your input will help create a comprehensive overview of common practices, tooling preferences, and challenges faced when deploying models to production, ultimately benefiting the entire ML community ?? We are also working on an interactive visualisation for everyone to be able to slice and dice across the data to derive meaningful insights on the production ML ecosystem!
The State of AI Report 2024 from AirStreet Capital is out! Really interesting insights from 2024, including highlights on the performance gap between proprietary models like GPT-4 and open-source alternatives has narrowed significantly, with advancements in planning, reasoning, and multimodal capabilities extending AI's applications into fields like mathematics, biology, and neuroscience. One interesting observation is that despite U.S. sanctions, Chinese labs continue to produce competitive AI models through alternative means. Similarly, the economic impact of global AI has surged, with public companies reaching a combined enterprise value of $9 trillion, although questions about long-term sustainability and viable business models remain. Check out the full insights on the report!
Uber has built a robust real-time data infrastructure to process Petabyte-scale data per day, using open-source technologies like Apache Kafka, Flink and Apache Pinot - these support production machine learning applications that require processing massive data volumes with low latency. Uber enables scalable stream processing and low-latency analytics critical for ML workflows like real-time prediction monitoring by enhancing these tools through leveraging frameworks such as FlinkSQL for easier streaming job creation with SQL and adding upsert capabilities to Pinot for real-time data updates. This is quite an interesting deep dive from Uber into their challenges and lessons learned throughout their large-scale data journey.
Microsoft has released an exciting project to bring GPU-native LLMs into CPUs with Bitnet.cpp; this is their official inference framework for 1-bit Large Language Models: This C++ library is inspired from Llama.cpp and provides optimized kernels for fast and energy-efficient inference on CPUs, with future support planned for NPUs and GPUs. This framework achieves significant speedups ?of 6×+ and energy reductions up to 82.2% on both ARM and x86 architectures, which is quite exciting for even 100B-parameter models.
Meta's FAIR team has released several new AI research artifacts to continue supporting the advancement of science across the community: These open source releases include: SAM 2.1 dataset on image and video segmentation; Meta Spirit LM as multimodal language model integrating speech and text; Layer Skip to accelerate LLM performance; SALSA to validate security for post-quantum cryptography standards; Meta Lingua for large-scale language model training; Meta Open Materials 2024 to accelerate AI-assisted inorganic materials discovery, and more.
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.
Software Engineer | Machine Learning | Full-Stack Developer | Flutter Developer | AWS | Project Manager | QML | Cloud Computing | CEO Codexbase Network | Manager SaaSSimplified |
1 个月very interesting i most say, can we attend this events online if i may ask?/