frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

Use "length of the RNN input" instead of "length of the RNN"

Browse Source 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/20873

Differential Revision: D15495570

Pulled By: ezyang

fbshipit-source-id: e3b4cd67ccf97d0053ac053c3bcb74415b928c0a
This commit is contained in:
Sergii Dymchenko
2019-05-24 09:00:26 -07:00
committed by Facebook Github Bot
parent 3e4f213e82
commit a5c90aaf47
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/source/notes/faq.rst
View File

@ -64,7 +64,7 @@ earlier, you should ``del intermediate`` when you are done with it.
**Don't run RNNs on sequences that are too large.** **Don't run RNNs on sequences that are too large.**
The amount of memory required to backpropagate through an RNN scales The amount of memory required to backpropagate through an RNN scales
linearly with the length of the RNN; thus, you will run out of memory linearly with the length of the RNN input; thus, you will run out of memory
if you try to feed an RNN a sequence that is too long. if you try to feed an RNN a sequence that is too long.
The technical term for this phenomenon is `backpropagation through time The technical term for this phenomenon is `backpropagation through time