Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Enable shared serialization in Torch #441

Merged
merged 1 commit into from
Nov 30, 2015
Merged

Enable shared serialization in Torch #441

merged 1 commit into from
Nov 30, 2015

Commits on Nov 29, 2015

  1. Enable shared serialization in Torch 

    This allows passing tensors between threads by
reference. This speeds up training for all models,
particulary when data are comparatively large
v.s. compute time.

Alexnet on CIFAR10 (2 epochs): 10m35s -> 9m28s
GoogleNet on CIFAR10 (1 epoch): 45m5s -> 44m52s
LeNet on MNIST (30 epochs): 2m17s -> 2m15s
    @gheinrich
    gheinrich committed Nov 29, 2015
    Configuration menu
    Copy the full SHA
    9bedefc View commit details
    Browse the repository at this point in the history