{ 'n_training_iterations': 10000, 'n_testing_iterations': 100, 'seed': 3000, 'cuda': False, 'rnn_type': <RNNType.GRU: 'gru'>, 'length': 2, 'width': 3, 'initial_delay': 1, 'initial_delay_fixed_length': True, 'delay': 2, 'delay_fixed_length': True, 'no_blank_symbol': False, 'batch_size': 1000, 'n_units': 50, 'learning_rate': 0.001, 'use_rmsprop': False, 'use_grad_clipping': False, 'grad_clip_norm': 2.0, 'binary_encoding': True, 'total_input_width': 5, 'total_input_length': 7, 'target_length': 2, 'target_width': 3, 'activity_regularization': False, 'activity_regularization_constant': 1.0, 'activity_regularization_target': 0.05 }
Seed: 3000 LSTM parameters: ['weight_ih_l0', 'weight_hh_l0', 'bias_ih_l0', 'bias_hh_l0'] Training iteration 0 :: Loss is 0.7040 :: Bitwise success rate 0.4290 (Running avg. 0.2145) :: Mean activity 0.0041 :: Batch time was 0.1784.
Training iteration 1 :: Loss is 0.7021 :: Bitwise success rate 0.4308 (Running avg. 0.3227) :: Mean activity 0.0033 :: Batch time was 0.1632.
Training iteration 100 :: Loss is 0.4692 :: Bitwise success rate 0.7607 (Running avg. 0.7597) :: Mean activity -0.0215 :: Batch time was 0.3571.
Training iteration 200 :: Loss is 0.3065 :: Bitwise success rate 0.8523 (Running avg. 0.8496) :: Mean activity -0.0230 :: Batch time was 0.1546.
Training iteration 273 :: Loss is 0.1485 :: Running avg. of bitwise success rate is high enough 0.9802. Stopping training.
Testing iteration 0 :: Loss is 0.1436 :: Mean activity -0.0390 :: Bitwise success rate 0.9815 :: Batch time was 0.1237. Testing iteration 1 :: Loss is 0.1470 :: Mean activity -0.0396 :: Bitwise success rate 0.9760 :: Batch time was 0.1237. Testing iteration 99 :: Loss is 0.1438 :: Mean activity -0.0393 :: Bitwise success rate 0.9802 :: Batch time was 0.1237.