运行出错，cuda10.0, torch 1.2.0

[Sohnia]

Traceback (most recent call last): File "train.py", line 250, in train(batchSize=8, device='cuda:0', lr=10e-4, use_dtp=True, pretrained_model=pretrained_model, tokenizer_model=tokenizer_model, weighted_loss=True) File "train.py", line 109, in train ocemotion_gold, tnews_kpi, ocnli_kpi, ocemotion_kpi) File "/home/zhugc19/tianchi-multi-task-nlp/calculate_loss.py", line 51, in compute_dtp if tnews_pred != None: TypeError: ne() received an invalid combination of arguments - got (NoneType), but expected one of:

• (Tensor other) didn't match because some of the arguments have invalid types: (NoneType)
• (Number other) didn't match because some of the arguments have invalid types: (NoneType)

[rogerrojur]

pytorch 版本 1.6.0 cuda 10.2 你看看readme