Default value of kernel_size

[SujanaC]

Describe the bug
It seems that TCN has a default kernel_size of 2 but the documentation indicates a value of 3. Could you please check this and fix the documentation?

Paste a snippet
from tcn import TCN, tcn_full_summary

time_steps = 10
input_dim = 2
tcn_layer = TCN(input_shape=(time_steps,input_dim), nb_filters=1, use_skip_connections=False)

tcn_layer.get_config()

Dependencies
I am running TF Keras 2.5.0 on a macbook pro with M1 chip.

[philipperemy]

Maybe update to the latest version?

{'name': 'tcn', 'trainable': True, 'batch_input_shape': (None, 10, 2), 'dtype': 'float32', 'nb_filters': 1, 'kernel_size': 3, 'nb_stacks': 1, 'dilations': (1, 2, 4, 8, 16, 32), 'padding': 'causal', 'use_skip_connections': False, 'dropout_rate': 0.0, 'return_sequences': False, 'activation': 'relu', 'use_batch_norm': False, 'use_layer_norm': False, 'use_weight_norm': False, 'kernel_initializer': 'he_normal'}

virtualenv venv
pip install keras-tcn
python test.py

Where test.py contains:

from tcn import TCN

time_steps = 10
input_dim = 2
tcn_layer = TCN(input_shape=(time_steps, input_dim), nb_filters=1, use_skip_connections=False)

print(tcn_layer.get_config())

I will close this issue. Let me know if you need some extra help.