未验证 提交 0b0d343b 编写于 作者: A asiacui 提交者: GitHub

change to Neural Recurrent Sequence Labeling Model for Open-Domain Factoid Question Answering_cn.md

上级 2dfab60c
......@@ -67,7 +67,7 @@
上述模型可以通过下面命令下载:
cd pre-trained-models && ./download-models.sh && cd ..
具有注释证据的测试集的评估结果可以利用下面的命令获取:
PYTHONPATH=data/evaluation:$PYTHONPATH python infer.py \
......@@ -97,4 +97,3 @@
--fuzzy --schema BIO2
# The result should be
# chunk_f1=0.749358 chunk_precision=0.727868 chunk_recall=0.772156 true_chunks=3024 result_chunks=3208 correct_chunks=2335
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册