accumulate_grad_batches:

1

min_epochs:

1

activation:

relu

model_depth:

50

alpha:

1

num_layers_CS:

2

batch_size:

16

num_layers_shared:

2

beta:

0.5

num_nodes:

1

check_val_every_n_epoch:

1

num_processes:

1

dropout:

0.5

num_sanity_val_steps:

2

gpus:

1

num_workers:

0

gradient_clip_val:

0

out_dim:

128

hidden_dim_CS:

128

overfit_batches:

0

hidden_dim_shared:

128

precision:

32

limit_test_batches:

1

process_position:

0

limit_train_batches:

1

progress_bar_refresh_rate:

0

limit_val_batches:

1

row_log_interval:

50

log_save_interval:

100

track_grad_norm:

−1

lr:

0.0003

val_check_interval:

1

max_epochs:

100

weight_decay:

1.00E−05