benchmark
7eb31b1d - Use the correct batch size for vgg16 training (#583)

Commit
4 years ago
Use the correct batch size for vgg16 training (#583) Summary: The batch size for train was set to 256. Source: https://arxiv.org/pdf/1409.1556.pdf Pull Request resolved: https://github.com/pytorch/benchmark/pull/583 Reviewed By: aaronenyeshi Differential Revision: D32841472 Pulled By: xuzhao9 fbshipit-source-id: b636a1fac01ce829b4b14834ee663f6ce7bd8d52
Author
Parents
Loading