These are chat archives for beniz/deepdetect
Thanks, I'll check it out. I tried to train vgg_face last night but verged on giving up.
Found a train_val + solver somewhere.. then issued the training call but kept getting a "Check failed (custom): (mdb_status) == (0)" error.
train/directory does not appear to hold the
1000classes you are declaring in the service creation.
lmdbfiles if any
db_lmdb.hpp:15 / Check failed (custom): (mdb_status) == (0)"
train_val.prototxtis wrong, you need to look at the
templates/caffe/vgg_16/vgg_16.prototxthead and fix it.
vgg_16.prototxt, that is not what you have here
./include/caffe/util/db_lmdb.hpp:15 / Check failed (custom): (mdb_status) == (0)"and
service face training call failedin the logs.