ljynlp/W2NER

IndexError

Fpw19980724 opened this issue · 5 comments

哈喽,我这里遇到这个问题,请问这是什么导致的?是数据量太大了吗?
Traceback (most recent call last):
File "main.py", line 276, in
datasets, ori_data = data_loader.load_data_bert(config)
File "W2NER-main/data_loader.py", line 205, in load_data_bert
train_dataset = RelationDataset(*process_bert(train_data, tokenizer, vocab))
File "W2NER-main/data_loader.py", line 144, in process_bert
_dist_inputs[i, j] = dis2idx[-_dist_inputs[i, j]] + 9
IndexError: index 1000 is out of bounds for axis 0 with size 1000

样本的长度太长了,可能超过了1000

好的感谢,那请问一下,如果训练完模型后单纯需要做预测,那么vocab那边初始化呀,因为我看代码里还是需要load training data, dev data 和 test data来初始化vocab的

utils.py中预留了save_fileload_file接口,可以通过这两个接口存储和加载vocab,具体可以根据自己的需求来实现