Pytorch running slower on GPU than CPU
up vote
0
down vote
favorite
I'm fairly new to python/machine learning and I've been trying to learn PyTorch. Here's a tutorial I've been following:
https://pytorch.org/tutorials/intermediate/char_rnn_classification_tutorial.html
, which is basically a very simple RNN to classify name into its language. When I run it on a CPU, the final results are
100000 100% (2m 25s) 0.3983 Tsujimoto / Japanese ✓
Training completed! Time taken: 145.90 seconds.
But when I run it on a GPU I get the follwing
100000 100% (5m 56s) 0.1462 Mcgregor / Scottish ✓
Training completed! Time taken: 356.96 seconds.
The only thing I changed was the train function by sending the input tensors to device (device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")) and of course putting the rnn itself to the GPU (rnn = RNN(n_letters, n_hidden, n_categories).to(device):
def train(category_tensor, line_tensor):
hidden = rnn.initHidden().to(device)
category_tensor, line_tensor = category_tensor.to(device), line_tensor.to(device)
rnn.zero_grad()
for i in range(line_tensor.size()[0]):
output, hidden = rnn(line_tensor[i], hidden)
loss = F.nll_loss(output, category_tensor)
loss.backward()
optimizer.step()
return output, loss.item()
Am I doing something wrong?
Thanks in advance!
python pytorch
asked Nov 20 at 8:47
Rcwt
1083
add a comment |
up vote
0
down vote
favorite
I'm fairly new to python/machine learning and I've been trying to learn PyTorch. Here's a tutorial I've been following:
https://pytorch.org/tutorials/intermediate/char_rnn_classification_tutorial.html
, which is basically a very simple RNN to classify name into its language. When I run it on a CPU, the final results are
100000 100% (2m 25s) 0.3983 Tsujimoto / Japanese ✓
Training completed! Time taken: 145.90 seconds.
But when I run it on a GPU I get the follwing
100000 100% (5m 56s) 0.1462 Mcgregor / Scottish ✓
Training completed! Time taken: 356.96 seconds.
The only thing I changed was the train function by sending the input tensors to device (device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")) and of course putting the rnn itself to the GPU (rnn = RNN(n_letters, n_hidden, n_categories).to(device):
def train(category_tensor, line_tensor):
hidden = rnn.initHidden().to(device)
category_tensor, line_tensor = category_tensor.to(device), line_tensor.to(device)
rnn.zero_grad()
for i in range(line_tensor.size()[0]):
output, hidden = rnn(line_tensor[i], hidden)
loss = F.nll_loss(output, category_tensor)
loss.backward()
optimizer.step()
return output, loss.item()
Am I doing something wrong?
Thanks in advance!
python pytorch
asked Nov 20 at 8:47
Rcwt
1083
2
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00
add a comment |
up vote
0
down vote
favorite
up vote
0
down vote
favorite
I'm fairly new to python/machine learning and I've been trying to learn PyTorch. Here's a tutorial I've been following:
https://pytorch.org/tutorials/intermediate/char_rnn_classification_tutorial.html
, which is basically a very simple RNN to classify name into its language. When I run it on a CPU, the final results are
100000 100% (2m 25s) 0.3983 Tsujimoto / Japanese ✓
Training completed! Time taken: 145.90 seconds.
But when I run it on a GPU I get the follwing
100000 100% (5m 56s) 0.1462 Mcgregor / Scottish ✓
Training completed! Time taken: 356.96 seconds.
The only thing I changed was the train function by sending the input tensors to device (device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")) and of course putting the rnn itself to the GPU (rnn = RNN(n_letters, n_hidden, n_categories).to(device):
def train(category_tensor, line_tensor):
hidden = rnn.initHidden().to(device)
category_tensor, line_tensor = category_tensor.to(device), line_tensor.to(device)
rnn.zero_grad()
for i in range(line_tensor.size()[0]):
output, hidden = rnn(line_tensor[i], hidden)
loss = F.nll_loss(output, category_tensor)
loss.backward()
optimizer.step()
return output, loss.item()
Am I doing something wrong?
Thanks in advance!
python pytorch
asked Nov 20 at 8:47
Rcwt
1083
I'm fairly new to python/machine learning and I've been trying to learn PyTorch. Here's a tutorial I've been following:
https://pytorch.org/tutorials/intermediate/char_rnn_classification_tutorial.html
, which is basically a very simple RNN to classify name into its language. When I run it on a CPU, the final results are
100000 100% (2m 25s) 0.3983 Tsujimoto / Japanese ✓
Training completed! Time taken: 145.90 seconds.
But when I run it on a GPU I get the follwing
100000 100% (5m 56s) 0.1462 Mcgregor / Scottish ✓
Training completed! Time taken: 356.96 seconds.
The only thing I changed was the train function by sending the input tensors to device (device = torch.device("cuda:3" if torch.cuda.is_available() else "cpu")) and of course putting the rnn itself to the GPU (rnn = RNN(n_letters, n_hidden, n_categories).to(device):
def train(category_tensor, line_tensor):
hidden = rnn.initHidden().to(device)
category_tensor, line_tensor = category_tensor.to(device), line_tensor.to(device)
rnn.zero_grad()
for i in range(line_tensor.size()[0]):
output, hidden = rnn(line_tensor[i], hidden)
loss = F.nll_loss(output, category_tensor)
loss.backward()
optimizer.step()
return output, loss.item()
Am I doing something wrong?
Thanks in advance!
python pytorch
python pytorch
asked Nov 20 at 8:47
Rcwt
1083
asked Nov 20 at 8:47
Rcwt
1083
asked Nov 20 at 8:47
Rcwt
1083
asked Nov 20 at 8:47
Rcwt
1083
asked Nov 20 at 8:47
Rcwt
1083
1083
2
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00
add a comment |
2
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00
2
2
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00
add a comment |
active
oldest
votes
Your Answer
StackExchange.ifUsing("editor", function () {
StackExchange.using("externalEditor", function () {
StackExchange.using("snippets", function () {
StackExchange.snippets.init();
});
});
}, "code-snippets");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "1"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53389218%2fpytorch-running-slower-on-gpu-than-cpu%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
active
oldest
votes
active
oldest
votes
active
oldest
votes
active
oldest
votes
Thanks for contributing an answer to Stack Overflow!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Some of your past answers have not been well-received, and you're in danger of being blocked from answering.
Please pay close attention to the following guidance:
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53389218%2fpytorch-running-slower-on-gpu-than-cpu%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
2
You may try different hidden_sizes, see here for this: stackoverflow.com/questions/53325418/… So with higher dim parameters it might be that GPU wins over CPU, however in general RNNs with for-loops are difficult to parallelize and can be painfully slow on GPU. So do not expect too much gain of it. To get most of RNNs it's best to use batching so that computation can be done in parallel.
– blue-phoenox
Nov 20 at 10:00