No description

katharinakorb c4a1737fa7 Add files via upload ... In der config Datei kann man nun einstellen, ob während des Training mit leaky relu, bei einer Performance von 100% auf relu geswitched wird (d.h. leaky relu mit slope = 0.0). In der cnn_trainin.py musste ich beim Lesen und Laden der config.json aufgrund eines komischen Errors beim Ausführen der .sh-file was ändern.		2023-08-01 10:16:27 +02:00
Classic_contour_net_shallow	Add files via upload	2023-08-01 10:16:27 +02:00
functions	Add files via upload	2023-07-29 03:23:24 +02:00
RenderStimuli	Add files via upload	2023-07-28 15:42:20 +02:00
clean.sh	Add files via upload	2023-07-28 01:19:55 +02:00
cnn_training.py	Add files via upload	2023-07-27 20:13:59 +02:00
cnn_training_v2.py	Add files via upload	2023-07-27 22:01:33 +02:00
config.json	Add files via upload	2023-07-23 14:07:20 +02:00
config_v2.json	Add files via upload	2023-07-27 22:01:33 +02:00
draw_kernels_v2.py	Add files via upload	2023-07-29 03:23:39 +02:00
inspect_weights_conv_0.py	Add files via upload	2023-07-20 01:51:27 +02:00
inspect_weights_conv_x.py	Add files via upload	2023-07-28 00:11:46 +02:00
LICENSE	Initial commit	2023-07-19 14:31:23 +02:00
network_0.json	Add files via upload	2023-07-28 00:11:46 +02:00
optimal_stimulus.py	Add files via upload	2023-07-29 16:01:51 +02:00
performance_pfinkel_plots.py	Add files via upload	2023-07-19 15:47:35 +02:00
plot_performance_data.py	Add files via upload	2023-07-19 18:11:50 +02:00
weights_0.npy	Add files via upload	2023-07-28 12:58:41 +02:00