Error in Classifier training

[ahsan856jalal]

@AlexeyAB
I have 15 class dataset and there is no problem in images and text files as I trained detector perfectly from the same dataset.
For training resnet50 classifier, my files are as following:

custom.labels.list

abudefduf vaigiensis
acanthurus nigrofuscus
amphiprion clarkii
chaetodon lununatus
chaetodon speculum
chaetodon trifascialis
chromis chrysura
dascyllus aruanus
dascyllus reticulatus
hemigumnus malapterus
myripristis kuntee
neoglyphidodon nigroris
pempheris vanicolensis
plectrogly-phidodon dickii
zebrasoma scopas

custom.shortnames.list

abudefduf vaigiensis
acanthurus nigrofuscus
amphiprion clarkii
chaetodon lununatus
chaetodon speculum
chaetodon trifascialis
chromis chrysura
dascyllus aruanus
dascyllus reticulatus
hemigumnus malapterus
myripristis kuntee
neoglyphidodon nigroris
pempheris vanicolensis
plectrogly-phidodon dickii
zebrasoma scopas

my .data file

classes=15
train  = /home/usr/data_train_hist_norm_640.list
valid  = /home/usr/data_train_hist_norm_640.list
backup = resnet_50
labels = cfg/custom.labels.list
names  = cfg/custom.shortnames.list
top=2

But I am getting this line:

Too many or too few labels: 0, /home/ahsanjalal/Fishclef/Datasets/Training_dataset/JPEGImages/00004861.jpg
2, 0.024: 0.000000, 0.000000 avg, 0.000000 rate, 3.289449 seconds, 128 images
Loaded: 0.000027 seconds

[ahsan856jalal]

Class names and folder names are same as in the custom.labels.list file

[Digital2Slave]

@ahsan856jalal same issue. Do u fix it ?

[AnhPC03]

You can reference here