search

Cartus / AGGCN

Attention Guided Graph Convolutional Networks for Relation Extraction (authors' PyTorch implementation for the ACL19 paper)
MIT License
433 stars 88 forks source link

关于准确率的问题 #6

Closed zhouyao1124 closed 5 years ago

zhouyao1124 commented 5 years ago

郭哥您好，在尝试复现论文代码时，发生了一些问题，执行语句为：python3 train.py --id $SAVE_ID --seed 0 --hidden_dim 300 --lr 0.7 --rnn_hidden 300 --num_epoch 100 --pooling max --mlp_layers 1 --num_layers 2 --pooling_l2 0.002

此时准确率均为100%且f1值均为0，详细截图如下： 2 1 希望您在方便时候能不吝赐教，谢谢！

Cartus commented 5 years ago

Hi,

I think the problem lies in the fact that the training and developing set is not complete. According to your last image, only 1 batch created for the training set.