Open modantailleur opened 1 month ago
& is shell token: https://www.oreilly.com/library/view/learning-the-bash/1565923472/ch01s09.html
I can try with ,
Here is a proposal:
python3 demo.py -f -s nn_type=cnn+lstm,dropout=0:nn_type=cnn,dropout=1 dropout=0+learning_rate=0.001+n_layers=2+nn_type=cnn dropout=0+learning_rate=0.0001+n_layers=2+nn_type=cnn dropout=0+learning_rate=0.00001+n_layers=2+nn_type=cnn dropout=0+learning_rate=0.001+n_layers=5+nn_type=cnn dropout=0+learning_rate=0.0001+n_layers=5+nn_type=cnn dropout=0+learning_rate=0.00001+n_layers=5+nn_type=cnn dropout=0+learning_rate=0.001+n_layers=8+nn_type=cnn dropout=0+learning_rate=0.0001+n_layers=8+nn_type=cnn dropout=0+learning_rate=0.00001+n_layers=8+nn_type=cnn dropout=0+learning_rate=0.001+n_layers=2+nn_type=lstm dropout=0+learning_rate=0.0001+n_layers=2+nn_type=lstm dropout=0+learning_rate=0.00001+n_layers=2+nn_type=lstm dropout=0+learning_rate=0.001+n_layers=5+nn_type=lstm dropout=0+learning_rate=0.0001+n_layers=5+nn_type=lstm dropout=0+learning_rate=0.00001+n_layers=5+nn_type=lstm dropout=0+learning_rate=0.001+n_layers=8+nn_type=lstm dropout=0+learning_rate=0.0001+n_layers=8+nn_type=lstm dropout=0+learning_rate=0.00001+n_layers=8+nn_type=lstm dropout=1+learning_rate=0.001+n_layers=2+nn_type=cnn dropout=1+learning_rate=0.0001+n_layers=2+nn_type=cnn dropout=1+learning_rate=0.00001+n_layers=2+nn_type=cnn dropout=1+learning_rate=0.001+n_layers=5+nn_type=cnn dropout=1+learning_rate=0.0001+n_layers=5+nn_type=cnn dropout=1+learning_rate=0.00001+n_layers=5+nn_type=cnn dropout=1+learning_rate=0.001+n_layers=8+nn_type=cnn dropout=1+learning_rate=0.0001+n_layers=8+nn_type=cnn dropout=1+learning_rate=0.00001+n_layers=8+nn_type=cnn
can you test and provide feedback ?
Could you allow to use a command like:
python main.py -s plan1/dataset=A&B+learning_rate=-3&-4
to select multiple factors in this non-dict selection mode ?