RAREMed
RAREMed copied to clipboard
Source code of our SIGIR'24 paper titled "Leave No Patient Behind: Enhancing Medication Recommendation for Rare Disease Patients".
作者你好,现在我按照readme里面操作之后,只要调整batch_size就会出现报错,应该是这段代码的问题: ``` def main_mask(): multi_target_dis = np.full((1, voc_size[0]), -1) multi_target_pro = np.full((1, voc_size[1]), -1) for i in range(batch_size): multi_target_dis[i, 0:len(batch[i][0])] = batch[i][0] multi_target_pro[i, 0:len(batch[i][1])] = batch[i][1] ``` 修改为 ``` actual_batch_size...
这会不会导致模型对NSP的遗忘问题?
DDI差距明显
你好,按你的readme中流程进行复现,其他指标都可以达到甚至超越论文中的效果,但DDI有明显差距,需要怎么调整? 