You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
One thing that was used extensively to help the training and the time it took to train the discocirc experiment was the saving and re-loading of previously computed tensornetwork contraction paths. See section 3.2.1 and Appendix G of the paper https://arxiv.org/abs/2409.08777. The contractor we used was the random greedy one from tensornetwork https://tensornetwork.readthedocs.io/en/latest/contractors.html.
The text was updated successfully, but these errors were encountered:
Request from @AnnaNPearson:
This is a request for a future release:
One thing that was used extensively to help the training and the time it took to train the discocirc experiment was the saving and re-loading of previously computed tensornetwork contraction paths. See section 3.2.1 and Appendix G of the paper https://arxiv.org/abs/2409.08777. The contractor we used was the random greedy one from tensornetwork https://tensornetwork.readthedocs.io/en/latest/contractors.html.
The text was updated successfully, but these errors were encountered: