f7e931ba3d
Additional files used for analysis |
||
---|---|---|
.. | ||
config.json | ||
get_trained_models.py | ||
pfinkel_performance_test64.py | ||
README.txt | ||
training_loop.sh |
Folder minimal_architecture: 1. config.json: * json file with all configurations and cnn parameters 2. training_loop.sh: * bash script to train the 64 cnns 3. get_trained_models: * searches for the saved trained models in a directory * chooses model based on the largest saved epoch in the save-name 4. pfinkel_performance_test64: * load all models extracted by 'get_trained_models' * test them on all stimulus conditions * sort their performances either after number of free parameters, or architecture