AaronHeee / Neural-Attentive-Item-Similarity-Model

TensorFlow Implementation of Neural Attentive Item Similarity Model for Recommendation on TKDE 2018
149 stars 48 forks source link

您好,执行报错,似乎是全局变量在多线程下为None,有什么方法能加个锁控制一下呢,谢谢。 #4

Open youth95 opened 5 years ago

youth95 commented 5 years ago

报错信息 Traceback (most recent call last): File "FISM.py", line 212, in training(args.pretrain, model, dataset, args.epochs, args.num_neg) File "FISM.py", line 139, in training batches = data.shuffle(dataset, model.batch_choice, num_negatives) File "C:\projects\Neural-Attentive-Item-Similarity-Model-master\Batch_gen.py", line 33, in shuffle return _preprocess(_get_train_batch_user) File "C:\projects\Neural-Attentive-Item-Similarity-Model-master\Batch_gen.py", line 64, in _preprocess res = pool.map(get_train_batch, range(_num_batch)) File "C:\ProgramData\Anaconda3\lib\multiprocessing\pool.py", line 266, in map return self._map_async(func, iterable, mapstar, chunksize).get() File "C:\ProgramData\Anaconda3\lib\multiprocessing\pool.py", line 644, in get raise self._value

muyunping123 commented 1 year ago

请问一下您是怎么解决这个问题的呢