hustcxx

bug：总是提示KeyError: 'sh'

2

@brxx122 您好： 下载了您提供的cnn的数据集，采用运行命令 python train.py --cuda --gpu 0 --data_dir ./data/middledata_2/ --cache_dir ./cache/cnn --embedding_path ./embedding_dir/glove.42B.300d.txt --model HSG --save_root ./data/model_path --log_root ./log --lr_descent --grad_clip -m 3，其他的都没有改过，但是提示sh的keyerror错误，找了很久没有发现错误是什么原因导致的，是否可以帮忙解答下。报错详细信息如下： result = self.forward(*input, **kwargs) File "/data/cxx/program/extractivemethod/heterogeneousgraph/module/GATLayer.py",...

in function get_state (),return Variable(torch.zeros(bsz, self.embedding_dim)).cuda(),and then is used in forward function of prob = self.softmax(torch.sum(m_A*u_temp, 2)),isn't the result zero???

some question about list_encode?

2

def forward(self,batch,pad=True): batch,phlens,batch_lens = batch batch_lens = tuple(batch_lens.tolist())#list-to-tuple _,enc = self.seqenc((batch,phlens)) enc = enc[:,2:] enc = torch.cat([enc[:,i] for i in range(enc.size(1))],1) m = max(batch_lens) encs = [self.pad(x,m) for x in...

Great jobs. I have some questions for the authors. 1. I run the code on the GPT-4 with the same parameter settings, but the results (macro-F1) for using GPT-4 as...