Can you provide a way to prepare the env?

[rayguan97]

1. Have you seen this error before? Could you give some insight on what could happened? """ cmd_str: cp ./models/pptnet.py exp/pptnet Traceback (most recent call last): File "train.py", line 709, in main() File "train.py", line 234, in main train(starting_epoch, model, optimizer, None, loss_function, lr_scheduler) File "train.py", line 241, in train train_one_epoch(model, optimizer, train_writer, loss_function, epoch) File "train.py", line 377, in train_one_epoch loss.backward() File "/home/anaconda3/envs/pptnet/lib/python3.6/site-packages/torch/tensor.py", line 195, in bac kward torch.autograd.backward(self, gradient, retain_graph, create_graph) File "/home/anaconda3/envs/pptnet/lib/python3.6/site-packages/torch/autograd/init.py", line 99, in backward allow_unreachable=True) # allow_unreachable flag RuntimeError: Expected isFloatingType(grads[i].type().scalarType()) to be true, but got false. (Could thi s error message be improved? If so, please report an enhancement request to PyTorch.)

"""

2. Can you provide the original environment for this code? I can't have this code to work after tired in several environment.