CUDA out of memory. Tried to allocate 20.00 MiB

[ForgetThatNight]

Traceback (most recent call last):
File "train.py", line 164, in
main()
File "train.py", line 160, in main
train_net(args)
File "train.py", line 67, in train_net
logger=logger)
File "train.py", line 111, in train
y_pred = model(x)
File "/home/ml/anaconda3/envs/Tacotron2-Mandarin/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/ml/Tacotron2-Mandarin/models/models.py", line 503, in forward
encoder_outputs, mels, memory_lengths=text_lengths)
File "/home/ml/anaconda3/envs/Tacotron2-Mandarin/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/ml/Tacotron2-Mandarin/models/models.py", line 404, in forward
decoder_input)
File "/home/ml/Tacotron2-Mandarin/models/models.py", line 360, in decode
attention_weights_cat, self.mask)
File "/home/ml/anaconda3/envs/Tacotron2-Mandarin/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/ml/Tacotron2-Mandarin/models/models.py", line 78, in forward
attention_hidden_state, processed_memory, attention_weights_cat)
File "/home/ml/Tacotron2-Mandarin/models/models.py", line 61, in get_alignment_energies
processed_query + processed_attention_weights + processed_memory))
RuntimeError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 7.79 GiB total capacity; 6.85 GiB already allocated; 13.56 MiB free; 23.24 MiB cached)
2080super上报OOM了，请问怎么处理

[sunxiaolongsf]

你好，我也遇到了这个问题，你解决了么

[ForgetThatNight]

你好，我也遇到了这个问题，你解决了么

还没有，代码运行在服务器，没办法进行debug调试，应该是计算loss的代码哪里有问题

[foamliu]

我在P100上测试通过。你们用的GPU是多大显存？把batch-size减半试试看？在utils.py文件中改。

[ForgetThatNight]

P100

我在P100上测试通过。你们用的GPU是多大显存？把batch-size减半试试看？在utils.py文件中改。

谢谢回答，是可以通过的， 看来 2080super 训练效果不会太好