search

flennerhag / warpgrad

Meta-Learning with Warped Gradient Descent
https://openreview.net/forum?id=rkeiQlBFPB
Apache License 2.0
92 stars 18 forks source link

About reproducing Table 4 in Appendix G #7

Closed haebeom-lee closed 3 years ago

haebeom-lee commented 3 years ago

Hi Flannerhag,

Congratulations for the nice work! I really enjoyed reading your paper.

Just a simple question about reproducing Table 4 in Appendix G. Could you let me know the reference code you used for KFAC? Or, do you have any plan to release your code for KFAC or any other baseline experiments based on random initializations?

And also, could you provide a brief explanation about to how use run_multi.py?

Thank you!

Best regards, Hae Beom Lee.

flennerhag commented 3 years ago

Hi,

Glad to hear you liked it! : )

Nowadays there are quite a few repo's out there, I'd expect any of them to reproduce (under the same hyper-params). For the paper, we used this one by the original authors of E-KFAC.

As for run_multi.py, it basically lets you exectute the main.py script for a given model over a range of pre-training tasks, for a given number of seeds per tasks. You can load and inspect the results in a notebook using the monitor.FileHandler class.