Fewshot-Learning-with-BERT question about label for calculating accuracy

question about label for calculating accuracy

Open zgf0907 opened this issue 11 months ago • 0 comments

train.py程序中计算准确率使用的label是随机生成的。个人认为应该调用模型生成的label与真实label进行比较，而非随机生成的label与真实label比较，该问题导致模型训练准确率一直在0.5左右 label = torch.arange(5).repeat(10).type(torch.LongTensor).cuda() correct += float(torch.sum(logits == label).item())

Aug 05 '23 08:08 zgf0907

Fewshot-Learning-with-BERT Fewshot-Learning-with-BERT copied to clipboard

question about label for calculating accuracy

Fewshot-Learning-with-BERT
Fewshot-Learning-with-BERT copied to clipboard