templateNER icon indicating copy to clipboard operation
templateNER copied to clipboard

Published 20 hours ago • verified publisher icon Nealcly

Cross-domain Few-Shot NER Result

Open lvminghan1997 opened this issue 2 years ago • 5 comments

in no source-domain data，MIT Movie 10 shot ,the result in paper is 37.3,why i get 51.06 , can you tell me some detail about that?

lvminghan1997 avatar May 04 '22 03:05 lvminghan1997

I have the same doubt, i get 0.496 in MIT Movie 10 shot

lplping avatar May 09 '22 07:05 lplping

I have the same doubt, i get 0.496 in MIT Movie 10 shot

您好！我现在已经在CoNLL03上复现了你的结果，也生成了对应的模型，请问我如何在MIT Movie少样本数据集上进行微调呢，直接在train.py中修改数据集的路径，然后运行就可以了吗?

zhanghaok avatar May 13 '22 10:05 zhanghaok

您好！我现在已经在CoNLL03上复现了你的结果，也生成了对应的模型，请问我如何在MIT Movie少样本数据集上进行微调呢，直接在train.py中修改数据集的路径，然后运行就可以了吗?

zhanghaok avatar May 13 '22 10:05 zhanghaok

One possible reason is the sample details. Could you sample multiple times and calculate the average performance?

Nealcly avatar May 16 '22 05:05 Nealcly

in no source-domain data，MIT Movie 10 shot ,the result in paper is 37.3,why i get 51.06 , can you tell me some detail about that?

Hello, I would like to ask if dev .txt is not needed in cross-domain experiments such as mit-movie. If not, will this overfit?

bwl666 avatar Apr 30 '23 03:04 bwl666