Data Science on AWS Meetup: Learning Quantum Computing and Pruning Deep Learning Models to Save Cost скачать в хорошем качестве

Data Science on AWS Meetup: Learning Quantum Computing and Pruning Deep Learning Models to Save Cost

Talk #1: Learning Quantum Machines Abstract: The state of quantum computing today resembles in many ways the early days of machine learning. Near term quantum algorithms are based on the idea of differentiable programming and “trainable” quantum algorithms. Much like in machine learning, our theoretical understanding of the computational power of these algorithms is limited, however, researchers in academia and industry are starting to experiment with early stage quantum computers to explore a variety of use cases from optimization to computational chemistry. In this talk you will learn how Amazon Braket allows customers to explore the potential of quantum computing, and how AWS is bringing mature ML tooling to quantum computing through the open source library PennyLane to help the two fields learn from each other and accelerate innovation. Speaker: Eric Kessler, PhD (  / eric-kessler-aws  ) Talk #2: Saving cost by pruning state-of-the-art deep learning models with Amazon SageMaker Debugger (https://www.amazon.science/blog/the-s...) Abstract: State-of-the-art deep learning models consist of millions of parameters and training such models from scratch is computationally intensive and can take hours, days, or even weeks. Especially in case of transfer learning, where one fine-tunes a pretrained model on a smaller dataset, one may not need a large number of parameters in the model - a smaller model may just perform as well. Model pruning can significantly reduce model size without sacrificing accuracy. The idea is simple: identify the redundant parameters in the model that contribute little to the training process. This talk will show how you can use Amazon SageMaker Debugger to identify the least important weights during training and then prune the model. Speakers: Nathalie Rauschmayr, Applied Scientist @ AWS and Connor Goggins, ML Engineer @ AWS RSVP Webinar: https://www.eventbrite.com/e/1-hr-fre... Zoom link: https://us02web.zoom.us/j/82308186562 Meetup: https://meetup.datascienceonaws.com Related Links ============= O'Reilly Book: https://www.amazon.com/dp/1492079391/ Website: https://datascienceonaws.com Meetup: https://meetup.datascienceonaws.com GitHub Repo: https://github.com/data-science-on-aws/ YouTube: https://youtube.datascienceonaws.com Slideshare: https://slideshare.datascienceonaws.com Support: https://support.pipeline.ai