Presentation: Petastorm: Training and Evaluation of Deep Learning Models

Share this on:

Abstract

Petastorm is an open source data access library that enables single machine or distributed training and evaluation of deep learning models directly from datasets in Apache Parquet format. Petastorm supports popular Python-based machine learning frameworks such as Tensorflow, Pytorch, and PySpark. In this talk, we present Petastorm features and show how Petastorm shortens the model development cycle at Uber ATG.

Speaker: Yevgeni Litvin

Tech Lead @Uber

Yevgeni is a tech lead on the Uber ATG Perception Team working on Machine learning infrastructure for training and evaluation of deep neural networks. 

2019 Tracks

  • Groking Timeseries & Sequential Data

    Techniques, practices, and approaches around time series and sequential data. Expect topics including image recognition, NLP/NLU, preprocess, & crunching of related algorithms.

  • Deep Learning in Practice

    Deep learning use cases around edge computing, deep learning for search, explainability, fairness, and perception.