def get_cfg(key=None):
if key is None:
cfg = CfgNode()
cfg.agg_type = 'concat'
# Attention Aggregator
cfg.attention_net = CfgNode()
cfg.attention_net.num_layers = 6
cfg.attention_net.units = 64
cfg.attention_net.num_heads = 4
cfg.attention_net.hidden_size = -1 # Size of the FFN network used in attention
cfg.attention_net.activation = 'gelu' # Activation of the attention
# Other parameters
cfg.mid_units = 128
cfg.feature_proj_num_layers = -1
cfg.out_proj_num_layers = 1
cfg.data_dropout = False
cfg.dropout = 0.1
cfg.activation = 'leaky'
cfg.normalization = 'layer_norm'
cfg.norm_eps = 1e-5
cfg.initializer = CfgNode()
cfg.initializer.weight = ['xavier', 'uniform', 'avg', 3.0]
cfg.initializer.bias = ['zeros']
else:
raise NotImplementedError
return cfg
def get_cfg(key=None):
if key is None:
cfg = CfgNode()
cfg.agg_type = 'concat'
cfg.mid_units = -1
cfg.num_layers = 0
cfg.data_dropout = False
cfg.dropout = 0.1
cfg.activation = 'tanh'
cfg.normalization = 'layer_norm'
cfg.norm_eps = 1e-5
cfg.initializer = CfgNode()
cfg.initializer.weight = ['xavier', 'uniform', 'avg', 3.0]
cfg.initializer.bias = ['zeros']
else:
raise NotImplementedError
return cfg