This week we celebrate breaking through 60,000 subscribers for the newsletter! Thank you to everyone for your continued support! |
|
|
|
|---|
|
|
MLL Sampling in Large Lang Models Chip Huyen dives into solutions for the probabilistic undeterminism of Large Language Models through robust sampling methods: In order to address the probabilistic nature of machine learning models in text generation we have to focusing on the challenges and strategies for achieving desired outputs. Sampling methods can be explored to address this, such as top-k, temperature and top-p sampling, highlighting their impact on the balance between creativity and consistency in model responses. The article also discusses test time sampling, a technique for improving model performance by generating multiple outputs and selecting the best one, and the concept of structured outputs, emphasizing the importance of guiding models to produce specific formats, especially in applications requiring precise structures like text-to-SQL. This piece is particularly insightful for ML practitioners looking to optimize text generation models in production environments. |
|
|---|
|
UK Govt's GenAI Framework The UK government releases their framework (and principles) for generative AI use, quite a lot of insightful takeaways from a public sector perspective: The UK Govt's GenAI framework was put together with a particular focus on Large Language Models, aiming to provide guidance for public sector applications. It emphasizes the importance of understanding AI's capabilities and limitations, advocating for lawful, ethical, and responsible usage - this is backed with sound deep dives for each of these topics. The framework outlines ten principles covering aspects such as security, human oversight, lifecycle management, and skill development. It addresses challenges like accuracy, bias, and environmental impacts, highlighting the need for transparency and meaningful human control. |
|
|---|
|
Why Machine Learning is Hard Why is machine learning hard? The machine learning field is considered 'hard' not only due to the scientific concepts involved, but due to the complexity in the workflows involved in the ML lifecycle, together with the intuition needed to select appropriate models and algorithms. Debugging in ML is uniquely challenging as it involves four dimensions: algorithm, implementation, data, and the model itself, making it more complex than traditional software debugging. This complexity is compounded by long feedback loops in training models, necessitating parallel experimentation and hindering sequential knowledge building. The core skill in ML, therefore, lies in developing an intuition for diagnosing and solving problems across these multiple dimensions, a skill honed through continuous practice and experience. |
|
|---|
|
Hand-picked AI Fundamentals An extensive and comprehensive hand-picked list of AI Fundamentals: The article provides a comprehensive guide for machine learning practitioners, encompassing a wide array of topics crucial to AI and ML. It covers design and architecture, including analysis of ML algorithms and training methodologies. The guide also delves into speech, vision and NLP (together with models like BERT and GPT). It covers multimodality, as well as privacy-preserving AI, evaluation techniques, MLOps considerations, miscellaneous foundational concepts, hyperparameter management, together with practice interview questions. |
|
|---|
|
Building a Brag-List for Success Top life-hack, keep a brag list of the wins you achieved: A great call to action for professionals in tech (+machine learning) fields, to maintain a "brag list" of your accomplishments. This enables you to effectively presenting your achievements during career advancements and helps you fight the annoying impostor syndrome by reminding yourself of your past successes. This is a technique that many practitioners (including myself) follow, and this has positive impact not only in career progression but even when colloquially discussing specific achievements. This article offers a great template to assist you in tracking your achievements. |
|
|---|
|
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 MLOps conferences:
In case you missed our key talks:
|
|
|---|
| | |
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. - Kompute - Blazing fast, lightweight and mobile phone-enabled GPU compute framework optimized for advanced data processing usecases.
- CuPy - An implementation of NumPy-compatible multi-dimensional array on CUDA. CuPy consists of the core multi-dimensional array class, cupy.ndarray, and many functions on it.
- Jax - Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
- CuDF - Built based on the Apache Arrow columnar memory format, cuDF is a GPU DataFrame library for loading, joining, aggregating, filtering, and otherwise manipulating data.
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! |
|
|---|
| | |
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: - MLSecOps Top 10 Vulnerabilities - This is an initiative that aims to further the field of machine learning security by identifying the top 10 most common vulnerabiliites in the machine learning lifecycle as well as best practices.
- AI & Machine Learning 8 principles for Responsible ML - The Institute for Ethical AI & Machine Learning has put together 8 principles for responsible machine learning that are to be adopted by individuals and delivery teams designing, building and operating machine learning systems.
- An Evaluation of Guidelines - The Ethics of Ethics; A research paper that analyses multiple Ethics principles.
- ACM's Code of Ethics and Professional Conduct - This is the code of ethics that has been put together in 1992 by the Association for Computer Machinery and updated in 2018.
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!
|
|
|---|
| | |
| | | | The Institute for Ethical AI & Machine Learning is a European research centre that carries out world-class research into responsible machine learning. | | | | |
|
|
|---|
|
|
This email was sent to You received this email because you are registered with The Institute for Ethical AI & Machine Learning's newsletter "The Machine Learning Engineer"
|
| | | | |
|
|
|---|
|
© 2023 The Institute for Ethical AI & Machine Learning |
|
|---|
|
|
|
The Institute for Ethical AI & Machine Learning
