cirrus

Serverless ML Framework
授权协议 Apache-2.0 License
开发语言 Python
所属分类 神经网络/人工智能、 机器学习/深度学习
软件类型 开源软件
地区 不详
投 递 者 公西繁
操作系统 跨平台
开源组织
适用人群 未知
 软件概览

Cirrus

Cirrus is a serverless machine learning library. Cirrus provides a list of machine learning algorithms that can scale to many serverless lambdas in the cloud.

Requirements

The Cirrus backend has been tested on Ubuntu 14.04/16.04/18.04 and Amazon AMI.

It has been tested with the following environment / dependencies:

  • g++-7

In the Amazon AMI please do:

$ sudo yum install glibc-static
$ sudo yum install openssl-static.x86_64
$ sudo yum install zlib-static.x86_64

In Ubuntu please do:

$ sudo apt-get install build-essential cmake automake zlib1g-dev libssl-dev libcurl4-nss-dev bison libldap2-dev libkrb5-dev

Building

$ ./bootstrap.sh
$ make -j 10

Paper

This project is part of a research project on Serverless Machine Learning Workflows. This works has been published and can be found here:

Joao Carreira, Pedro Fonseca, Alexey Tumanov, Andrew Zhang, Randy Katz.In the ACM Symposium on Cloud Computing 2019 (SoCC'19)

Funding

This work has been generously supported by AWS Cloud Research, FCT (Portuguese Science Foundation), NSF CISE Expeditions Award CCF-1730628, and gifts from Alibaba, Amazon Web Services, Ant Financial, CapitalOne, Ericsson, Facebook, Futurewei, Google, Intel, Microsoft, Nvidia, Scotiabank, Splunk and VMware.

Contributors

Joao Carreira, Andrew Zhang, Jeff Yu, Ryan Wang, Neel Somani, Shea Conlon, Andy Wang, Pedro Fonseca, Alexey Tumanov, Randy Katz

  • 描述Cirrus logic 音频放大器CS35L45-CWZRFAE:13723714318 CS35L45 是业界最先进的智能升压放大器解决方案,可在不影响回放音频质量的情况下,实现移动设备最大音量。该集成电路 (IC) 的硬件平台结合了 6.8 W D 级单声道放大器,具有同类放大器中最低的噪声和静态功耗。数字控制的 H 级感应升压转换器能够将电池电压提高至 15 V,以进行扬声器回放。 扬

  • CS5343/4是用于数字音频系统的完整数模转换器。它执行采样、模数转换和抗混叠滤波,以串行形式为左输入和右输入生成24位值,每个通道的采样率高达108 kHz。 CS5343/4使用三阶多位Delta Sigma调制器,后跟数字滤波器,无需外部抗混叠滤波器。 CS5343/4还具有高阻抗采样网络,消除了运算放大器等昂贵的外部组件。 CS5343/4采用10针TSSOP封装,适用于商用(-40°至