Arjun Krishnakumar comments

Results 9 comments of


                                            Arjun Krishnakumar

How to recover the search space from `NAS-Bench-301`?

Hi @nzw0301, Thank you for your interest in our project and sorry for the delay in replying. The conversion to the NASLib graph can be found here: https://github.com/automl/NASLib/blob/1dd550774821799405e9a8295d4d7638e49983dc/naslib/search_spaces/darts/conversions.py#L310 I hope...

How to recover the search space from `NAS-Bench-301`?

Hi @nzw0301, it looks like this is due to an update in the latest version of `networkx`. Could you uninstall it and install version 2.8? I've also updated `requirements.txt`. You...

Extend predictors to zero-cost case

Ideally, yes. But there are about 19 predictors in the original ensemble. Let's focus now on extending the ZC case to the tree-based predictors, which are `LGBoost`, `NGBoost`, and `RandomForestPredictor`....

Extend predictors to zero-cost case

Best to be able to read from config file, too.

Storing of models or architectures in optimizers

We used only random sampling for the optimization in the paper. You can change the mutation code to comply with the RS acquisition function.

Source of validation accuracy in zero-cost case

The code was written this way for the Zero-Cost NAS paper, where we consumed only search spaces for which the values were available in the zc_api. It would make more...

Checkpoint restart training

Hi @Mars-204, Thanks for bringing this to our attention! Could you please tell us which search space you're using? Best, Arjun

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.DoubleTensor [1, 512, 8, 8]], which is output 0 of ReluBackward0, is at version 1; expected version 0 instead.

Hi @thompsondd, Could you please tell us which proxy you were using? Looks to me like removing an inplace relu operation somewhere in the Nasbench101 graph will fix the issue....

Val_accuracy is negative for "room_layout " in "zc_transbench101_micro.json" and "zc_transbench101_macro.json"

Hi @SliverySky , Thank you for your interest in our work! The `room-layout` task of Trans-NAS-Bench-101 is to estimate a 3d bounding box, and as such, it does not have...