insightface Subcenter ArcFace drop.py

@nttstar I checked the pre-trained model that was made available and there is no fc7_*_weight layers - referred in Line 100 of drop.py (Using r100's pretrained models)I think these layers comes from parallal_local_module_v1? Is there any new push to be done to this repo? Trying to run drop.py -- any suggestions?

Aug 29 '22 14:08 abhatta1234

Hi, you have to train from scratch using ArcFace and also save the 'fc7_*' weights which is required by drop functions.

Aug 29 '22 16:08 nttstar

@nttstar Thank you for your suggestion! I did the training part but couldn't save "fc7_" weights. Any suggestions? I guess I am confused about saving "fc7_" for label_name in model/while calling parallel module. I can't find any documentation to be able to save fc7 weights. Thank you for your help!!

Aug 29 '22 20:08 abhatta1234

set config.ckpt_embedding = False may work.

Aug 30 '22 01:08 nttstar

@nttstar Did that too, but it seems to be not working. When I get parallel module returned, it seems like it's not returning fc7_* layers. Seems like something is missing with either self.params_initialized or set_params(). Seems like set_params() adds fc7_* to parallel modules. Any recommendations??

Aug 30 '22 16:08 abhatta1234

@abhatta1234 I am facing same issue, were you able to solve this?

Jul 11 '23 06:07 PR451