This is the model in "A Graphical and Attentional Framework for Dual-Target Cross-Domain Recommendation" (IJCAI2020). GA-DTCDR is an optimized model for DTCDR ("DTCDR: A Framework for Dual-Target Cross-Domain Recommendation" in CIKM2019).
I ran the dataset douban_book_music, encoutering the problem that the nan loss and the test metric is fixed. Would you kindly tell how to solve it, e.g., tuning what the parameters.
I ran the dataset
douban_book_music, encoutering the problem that the nan loss and the test metric is fixed. Would you kindly tell how to solve it, e.g., tuning what the parameters.Experimental Results Log.
Start Evaluation! Epoch 279 Domain A: douban_book TopK: 10 HR: 0.34828009828009826, NDCG: 0.19299960999304305 Epoch 279 Domain B: douban_music TopK: 10 HR: 0.2963302752293578, NDCG: 0.16454914953441419 ====================Epoch 279 End==================== ====================Epoch 280 ==================== 170 / 178 : loss = nan;1076660156255; Mean loss in this epoch is: Domain A=nan;Domain B=nan ================================================== Start Evaluation! Epoch 280 Domain A: douban_book TopK: 10 HR: 1.0, NDCG: 0.33333333333333337 Epoch 280 Domain B: douban_music TopK: 10 HR: 1.0, NDCG: 0.3333333333333334