dalinvip / pytorch_joint-word-segmentation-and-pos-tagging Goto Github PK

小批量几百行自建数据集测试时可以通过，使用完整的自建数据集时就报错。
自建数据集部分格式如下：
話_NULL 說_NULL 天下_EPB 大勢_EPE ，_NN _PTA
分久必合_ID ，_PTA
合久必分_ID 。_PTA

RuntimeError: cuda runtime error (38) : no CUDA-capable device is detected at /opt/conda/conda-bld/pytorch_1549627089062/work/torch/csrc/cuda/Module.cpp:34

What is the meaning of pos and seg respectively when classifying

Traceback (most recent call last):
File "main.py", line 157, in
main()
File "main.py", line 74, in main
train_iter, dev_iter, test_iter, alphabet, alphabet_static = load_data(config=config)
File "/home/zgy/wll/tibetan_Segand-POS-Tag/DataUtils/mainHelp.py", line 242, in load_data
train_iter, dev_iter, test_iter, alphabet, alphabet_static = preprocessing(config)
File "/home/zgy/wll/tibetan_Segand-POS-Tag/DataUtils/mainHelp.py", line 168, in preprocessing
train_iter, dev_iter, test_iter = create_iter.createIterator()
File "/home/zgy/wll/tibetan_Segand-POS-Tag/DataUtils/Batch_Iterator.py", line 84, in createIterator
self._convert_word2id(self.data[id_data], self.operator, self.operator_static)
File "/home/zgy/wll/tibetan_Segand-POS-Tag/DataUtils/Batch_Iterator.py", line 112, in _convert_word2id
pos = inst.pos[index]
IndexError: list index out of range

我自己尝试的写了一下测试解码部分，还是有点问题，作者什么时候有时间更新一下测试解码部分