One of the best works for semantic seg with NAS. Thank You for sharing train as well as search code.
I downloaded student's weights from the location mentioned in the readme https://drive.google.com/file/d/1O56HnA0ug2M3K4SR3_AUzIs0wegy9BX6/view?usp=sharing and ran train.py to check the validation accuracy but I got 70.5%.
Is this model supposed to generate 70.5% or 73.1% val accuracy?
If there is some other weights for getting 73.1% can you pls share it?
One of the best works for semantic seg with NAS. Thank You for sharing train as well as search code.
I downloaded student's weights from the location mentioned in the readme https://drive.google.com/file/d/1O56HnA0ug2M3K4SR3_AUzIs0wegy9BX6/view?usp=sharing and ran train.py to check the validation accuracy but I got 70.5%. Is this model supposed to generate 70.5% or 73.1% val accuracy? If there is some other weights for getting 73.1% can you pls share it?
Thanks.