Serial computation: In 2019, how many minutes will it take to train ResNet-152 on ILSVRC 2012 using a system available on the public market for less than $5000?

Your submission is now in Draft mode. Once it's ready, please submit your draft for review by our team of Community Moderators. Thank you!

Question

This question is closely related to the ResNet-152 Parallel computation question.


The ImageNet Large Scale Visual Recognition Challenge consists in classifying objects in images. It ran from 2010 to 2017 (and then morphed into an object localization challenge living indefinitely on Kaggle).

It was in the 2012 challenge that the deep convolutional network AlexNet made a dramatic breakthrough (achieving an error rate around ~16%, compared to the previous record of ~25%), an event widely considered to have launched the deep learning revolution of the 2010s.

A second major milestone was the introduction of residual networks (ResNets) by Microsoft researchers in 2015, including the 152-layered ResNet-152 used as a benchmark for this question.

In particular, we are referring to the ResNet-152 model executed in this TensorFlow benchmark, trained until it reaches top-1 error of <=28% and top-5 error <7% (these numbers indicates whether the score only counts the network's single best guess for the image label, or allows it to provide its top 5 best guesses).

Unless an experiment as outlined above is actually run, the question will resolve as ambiguous. Please condition your forecasts on that assumption.


Data:

Previous ILSVRC results can be found on image-net.org/challenges/LSVRC/201X/results by substituting X for the relevant year.

EFF summarises performance of the winning algorithm from each year's challenges. (Note that the dataset was significantly expanded in 2014, and possibly in other years, whereas this question refers specifically to the 2012 dataset.)

This page summarises the performance and training time of various models on ILSVRC 2012, using various GPUs. (Note that these models were written in Torch, whereas the benchmark referred in the question was written in Tensorflow.)

Make a Prediction

Prediction

Note: this question resolved before its original close time. All of your predictions came after the resolution, so you did not gain (or lose) any points for it.

Note: this question resolved before its original close time. You earned points up until the question resolution, but not afterwards.

This question is not yet open for predictions.

Current points depend on your prediction, the community's prediction, and the result. Your total earned points are averaged over the lifetime of the question, so predict early to get as many points as possible! See the FAQ.

Metaculus help: Predicting

Predictions are the heart of Metaculus. Predicting is how you contribute to the wisdom of the crowd, and how you earn points and build up your personal Metaculus track record.

The basics of predicting are very simple: move the slider to best match the likelihood of the outcome, and click predict. You can predict as often as you want, and you're encouraged to change your mind when new information becomes available.

The displayed score is split into current points and total points. Current points show how much your prediction is worth now, whereas total points show the combined worth of all of your predictions over the lifetime of the question. The scoring details are available on the FAQ.

Thanks for predicting!

Your prediction has been recorded anonymously.

Want to track your predictions, earn points, and hone your forecasting skills? Create an account today!

Track your predictions
Continue exploring the site