TF-NAS icon indicating copy to clipboard operation
TF-NAS copied to clipboard

TF-NAS: Rethinking Three Search Freedoms of Latency-Constrained Differentiable Neural Architecture Search (ECCV2020)

Results 3 TF-NAS issues
Sort by recently updated
recently updated
newest added

Hi @AberHu, Thank you for uploading your code. It would be great if you provide comments on what each variable means in the `make_lat_lut_example.py` file.

In [train_search.py](https://github.com/AberHu/TF-NAS/blob/35a34a11b6a64ecf1047cb7acb016e04f99ea259/train_search.py#L422), I noticed that you do log_softmax() out of the graph, but why? Why not just use param **alpha** instead and do log_softmax() in each forward step?