📝 Editorial 

TensorFlow and PyTorch have become the two most popular deep learning frameworks within the data science community. Their dominance in the market is strong. And it increases as both frameworks have been very fast to incorporate cutting-edge deep learning methods and ML engineering techniques that can help accelerate the implementation of deep learning solutions. As a result, each release of these frameworks drives a lot of attention across the data science space. Last week was TensorFlow’s turn with the release of its 2.10 version.  

TensorFlow 2.10 has plenty of interesting features, but none were more notable than the improvements in Keras’ capabilities. The new version of TensorFlow includes improves Keras’ attention layers with features such as causal attention and implicit masking. The release also includes a new Keras optimizer API and changes in the Keras initializers to make them both deterministic and stale. Outside Keras, TensorFlow 2.10 includes hardware optimizations such as the support for the Compute Library for the Arm® Architecture (ACL) as well as a wider GPU support for Windows. Other notable features include a stable release of the popular TensorFlow Decision Forests (TF-DF), including JavaScript and Go inference APIs.  

TensorFlow 2.10 shouldn’t be considered a major release but certainly incorporates many features that have been highly demanded by the developer community. An improved experience for Keras developers, wider hardware topology coverage, and improved libraries were at the center of this release. As usual, the frantic pace of innovation in TensorFlow and PyTorch does not disappoint.       

🔺🔻TheSequence Scope – our Sunday edition with the industry’s development overview – is free. To receive high-quality content about the most relevant developments in the ML world every Tuesday and Thursday, please subscribe to TheSequence Edge 🔺🔻

🗓 Next week in TheSequence Edge:

Edge#225: we explain latent diffusion models; discuss the original latent diffusion paper; explore Hugging Face Diffusers, a library for state-of-the-art diffusion models.

Edge#226: we deep dive DeepSpeed Compression, a new library for extreme compression of deep learning models

🗺 Help us shape the ML value chain landscape

Let’s create an objective landscape of the ML Value Chain together. You’ve probably seen some AI/ML companies’ landscapes before. They are typically assembled by either analyst firms, media, or VC firms. But we trust that TheSequence’s audience only can shape an accurate landscape of the ML Value Chain. Participate and be the first to receive this super helpful research shaped by you!

PARTICIPATE NOW

Now, let’s review the most important developments in the AI industry this week

🔎 ML Research

Conversational Agents and Human Values 

DeepMind published a fascinating research paper drawing lessons from philosophy and linguistics to improve the alignment of conversational agents and human values →read more 

Fairness in Connection Recommender Systems 

Researchers from Carnegie Mellon University (CMU) published a paper outlining fairness techniques for connection recommender systems which are common in social networking platforms →read more

Auto Scheduler Optimizations 

Amazon Research published a paper introducing DietCode, an auto-scheduler method that can drastically optimize tensor operations in deep learning models →read more 

Game Theory and Offline RL 

Microsoft Research published two papers detailing a game theoretic approach to improve offline reinforcement learning models →read more

Better Transformers for Computer Vision 

Google Research published a paper detailing a multi-axis technique to improve computer vision transformer architectures →read more

📌 Event: Learn strategies to scale your ML models using Kubernetes - SEP 14

Get ahead of the curve, and learn practical hands-on guidance from Kubernetes expert Itay Ariel on how to leverage Kubernetes for distributed workloads. Itay will give an overview of the unique challenges of scaling workloads and show how to leverage Kubernetes to easily scale your ML models and automate the management of workload performance.

REGISTER FOR FREE

🤖 Cool AI Tech Releases

TensorFlow 2.10 

A new version of TensorFlow has been related with an exciting set of capabilities →read more

MLPerf 

MLCommons unveiled a new series of MLPerf benchmarks that evaluate the performance of inference models →read more

🛠 Real World ML 

MLOps at Walmart  

Walmart discusses some insightful details about the MLOps infrastructure and processes powering their ML pipelines →read more

💸 Money in AI

ML

• Observability platform Arize AI raised $38 million in a Series B funding round led by TCV. Hiring remote.

AI-driven

• Conversational AI startup PolyAI raised $40 million in a Series B funding round led by Georgian. Hiring in New York/US and London/UK.

• Document search tools developer Hebbia raised $30 million in a Series A round led by Index Ventures. Hiring in New York/US.

• Web-3 application developer platform Spice AI raised $13.5 million in a seed round led by Madrona Venture Group. Hiring in Seattle/US.