[R] ResNet strikes back: An improved training procedure in timm. There has been significant progress on best practices for training neural nets since ResNet's introduction in 2015. With such advances, a vanilla ResNet-50 reaches 80.4% top-1 accuracy on ImageNet without extra data or distillation.

• assembled-cnn

  1 330 0.0 Python

  Tensorflow implementation of "Compounding the Performance Improvements of Assembled Techniques in a Convolutional Neural Network"

  

As far as i know, the assemble-ResNet-50 (https://github.com/clovaai/assembled-cnn) gets 82.8% top-1, though they make some (minor) changes to ResNet-50 architecture.

• InfluxDB

  www.influxdata.com featured

  Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.

  

Suggest a related project