search

stanfordnlp / wge

Workflow-Guided Exploration: sample-efficient RL agent for web tasks
https://stanfordnlp.github.io/wge/
Other
110 stars 33 forks source link

Only works with Pytorch<=0.1.12 #1

Closed muuoyaar closed 6 years ago

muuoyaar commented 6 years ago

First of all, many thanks for all the work you put into making sure your work is reproducible! 🥇

It seems that ever since Pytorch==0.2.0 it is assumed that keepdim=False for reduction functions such as tensor.sum(). This assumption breaks a bunch of the code here... so it may be wise to specify a version in the Readme.

ezliu commented 6 years ago

Thanks for pointing this out! I've fixed the README