low_level_inplanes is not match at decode of build autodeeplab #93

MasatakaShibuya · 2020-01-27T06:06:47Z

https://github.com/NoamRosenberg/AutoML/blob/new_master/retrain_model/build_autodeeplab.py#L25
filter_multiplier=args.filter_multiplier * args.block_multiplier

multiply filter_param_dict.
If low_level_feature is 3rd layer
https://github.com/NoamRosenberg/AutoML/blob/new_master/retrain_model/new_model.py#L158
->
filter_multiplier=args.filter_multiplier * args.block_multiplier * filter_param_dict[network_path[2]]

…repository.

ZhengPeng7 mentioned this issue Nov 27, 2020

RuntimeError: Given groups=1, weight of size 48 160 1 1, expected input[2, 640, 3, 3] to have 160 channels, but got 640 channels instead #100

Open

vdesgrange added a commit to vdesgrange/autodeeplab-1 that referenced this issue May 4, 2021

Try fix decoding related issue. See Issue NoamRosenberg#93 in github …

17f54cc

…repository.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

low_level_inplanes is not match at decode of build autodeeplab #93

low_level_inplanes is not match at decode of build autodeeplab #93

MasatakaShibuya commented Jan 27, 2020 •

edited

Loading

low_level_inplanes is not match at decode of build autodeeplab #93

low_level_inplanes is not match at decode of build autodeeplab #93

Comments

MasatakaShibuya commented Jan 27, 2020 • edited Loading

MasatakaShibuya commented Jan 27, 2020 •

edited

Loading