A Bug (maybe) in "utils.py"

[Looong01]

At the line 86, the bug should be corrected into this as follow:

def mix_iters(iters): table = [] for i, iter in enumerate(iters): table += [i] * len(iter) np.random.shuffle(table) for i in table: 86 ----> # yield iters[i].next() ----> yield next(iters[i])

When I try to start training, I encountered a error as it run into line 86. It returned that the iter has no attribute called next. And then I search all the things on Google or Bing, and I cannot find any useage like this "iter.next()". The unique method is "next(iter)". After that I change it and run it successfully.