AttributeError: 'ReLU' object has no attribute 'weight'

[dupanfei1]

Hi, I met a problem in CNNCharEmbedding.py. The ReLU has no weight. It seems like there should be a difference for the activation functions with weight and without weight.
Thank you !

Traceback (most recent call last):
File "train.py", line 372, in
main(params)
File "train.py", line 283, in main
lm = LearningMachine('train', conf, problem, vocab_info=vocab_info, initialize=initialize, use_gpu=conf.use_gpu)
File "/nb_code1/LearningMachine.py", line 38, in init
self.model = Model(conf, problem, vocab_info, use_gpu)
File "/nb_code1/Model.py", line 210, in init
self.add_layer(EMBED_LAYER_ID, get_layer(layer_arch['layer'], all_layer_configs[EMBED_LAYER_ID]))
File "/nb_code1/Model.py", line 121, in get_layer
layer = eval(layer_name)(conf)
File "/nb_code1/block_zoo/Embedding.py", line 125, in init
self.embeddings[input_cluster] = eval(layer_conf.conf[input_cluster]['type'])(char_emb_conf)
File "/nb_code1/block_zoo/embedding/CNNCharEmbedding.py", line 81, in init
self.activation.weight = torch.nn.Parameter(self.activation.weight.cuda())
File "anaconda3/envs/py36/lib/python3.6/site-packages/torch/nn/modules/module.py", line 535, in getattr
type(self).name, name))
AttributeError: 'ReLU' object has no attribute 'weight'

[adolphk-yk]

Yes, some activation function have no weight. We have fixed this problem in the new pr. Thanks.