-
Notifications
You must be signed in to change notification settings - Fork 23
deepbinner train
Ryan Wick edited this page Aug 20, 2018
·
1 revision
usage: deepbinner train [-h] --train TRAIN --val VAL --model_out MODEL_OUT
[--model_in MODEL_IN] [--epochs EPOCHS] [--aug AUG]
[--batch_size BATCH_SIZE]
[--batches_per_epoch BATCHES_PER_EPOCH]
Train the neural network
optional arguments:
-h, --help show this help message and exit
Required:
--train TRAIN Balanced training data produced by the balance command
--val VAL Validation data used to assess the training
--model_out MODEL_OUT Filename for the trained model
Other:
--model_in MODEL_IN An existing model to use as a starting point for
training
--epochs EPOCHS Number of training epochs (default: 100)
--aug AUG Data augmentation factor (1 = no augmentation)
(default: 2.0)
--batch_size BATCH_SIZE Training batch size (default: 20)
--batches_per_epoch BATCHES_PER_EPOCH
The number of samples per epoch will be this times the
batch size (default: 5000)