rikdz
commented
5 years ago Reduce batch sizes with -t1size -t2size -t3size flags
Open sjzabc opened 5 years ago
rikdz
commented
5 years ago Reduce batch sizes with -t1size -t2size -t3size flags
wutaiqiang
commented
4 years ago 推荐使用TiTAN X
~/桌面/GraphWriter-master$ python3.6 train.py -save res Save File Exists, OverWrite? for no
Loading Data from data/preprocessed.train.tsv
building vocab
done
Sorting training data by len
ds sizes: 11880 26156 684 1000 Vocab sizes:
src 6343
ent 53343
nerd 8
rel 17
out 11738
graph
cuda:0
epoch 0 lr 0.1
Training 1
/home/ryf/桌面/GraphWriter-master/models/last_graph.py:75: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requiresgrad(True), rather than torch.tensor(sourceTensor).
ngraph = torch.tensor(vgraph.repeat(N,1).view(N,N,-1),requires_grad=False)
/home/ryf/桌面/GraphWriter-master/models/newmodel.py:62: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requiresgrad(True), rather than torch.tensor(sourceTensor).
cx = torch.tensor(hx)
Traceback (most recent call last):
File "train.py", line 130, in
main(args)
File "train.py", line 115, in main
train(m,o,ds,args)
File "train.py", line 32, in train
p,z,planlogits = m(b)
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 547, in call
result = self.forward(*input, kwargs)
File "/home/ryf/桌面/GraphWriter-master/models/newmodel.py", line 96, in forward
_, z = self.mattn(l,(ents,entlens))
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 547, in call
result = self.forward(*input, *kwargs)
File "/home/ryf/桌面/GraphWriter-master/models/attention.py", line 15, in forward
decsmall = self.attnlin(dec)
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 547, in call
result = self.forward(input, kwargs)
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/linear.py", line 87, in forward
return F.linear(input, self.weight, self.bias)
File "/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py", line 1371, in linear
output = input.matmul(weight.t())
RuntimeError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 2.95 GiB total capacity; 2.41 GiB already allocated; 14.19 MiB free; 51.57 MiB cached)