我创建了一个火炬张量，但我希望它进入GPU，但不是。太坏了怎么了？
==================================================================================================== FAILURES =====================================================================================================
_____________________________________________________________________________________________ test_model_works_on_gpu _____________________________________________________________________________________________
def test_model_works_on_gpu():
with torch.cuda.device(0) as cuda:
some_random_d_model = 2 ** 9
five_sentences_of_twenty_words = torch.from_numpy(np.random.random((5, 20, T * d))).float()
five_sentences_of_twenty_words_mask = torch.from_numpy(np.ones((5, 1, 20))).float()
pytorch_model = make_sentence_model(d_model=some_random_d_model, T_sgnn=T, d_sgnn=d)
five_sentences_of_twenty_words.to(cuda)
five_sentences_of_twenty_words_mask.to(cuda)
print(type(five_sentences_of_twenty_words), type(five_sentences_of_twenty_words_mask))
print(five_sentences_of_twenty_words.is_cuda, five_sentences_of_twenty_words_mask.is_cuda)
pytorch_model.to(cuda)
output_before_match = pytorch_model(five_sentences_of_twenty_words, five_sentences_of_twenty_words_mask)
assert output_before_match.shape == (5, some_random_d_model)
print(type(output_before_match))
> print(output_before_match.is_cuda, output_before_match.get_device())
E RuntimeError: get_device is not implemented for tensors with CPU backend
tests/test_model.py:58: RuntimeError
---------------------------------------------------------------------------------------------- Captured stdout call -----------------------------------------------------------------------------------------------
<class 'torch.Tensor'> <class 'torch.Tensor'>
False False
<class 'torch.Tensor'>
也：
>>> torch.cuda.is_available()
True
>>> torch.cuda.device_count()
2
并且：
pip freeze | grep -i torch
torch==1.0.0
torchvision==0.2.1
答案 0 :(得分：1)
您的问题是以下几行：
five_sentences_of_twenty_words.to(cuda)
five_sentences_of_twenty_words_mask.to(cuda)
.to(device) 仅在应用于模型时就地运行。
当应用于张量时，它必须被赋值：
five_sentences_of_twenty_words = five_sentences_of_twenty_words.to(cuda)
five_sentences_of_twenty_words_mask = five_sentences_of_twenty_words_mask.to(cuda)
答案 1 :(得分：0)
此操作已解决：
def test_model_works_on_gpu():
device_id = 0
with torch.cuda.device(device_id) as cuda:
some_random_d_model = 2 ** 9
five_sentences_of_twenty_words = torch.from_numpy(np.random.random((5, 20, T * d))).float()
five_sentences_of_twenty_words_mask = torch.from_numpy(np.ones((5, 1, 20))).float()
pytorch_model = make_sentence_model(d_model=some_random_d_model, T_sgnn=T, d_sgnn=d)
five_sentences_of_twenty_words = five_sentences_of_twenty_words.cuda(device_id)
five_sentences_of_twenty_words_mask = five_sentences_of_twenty_words_mask.cuda(device_id)
print(type(five_sentences_of_twenty_words), type(five_sentences_of_twenty_words_mask))
print(five_sentences_of_twenty_words.is_cuda, five_sentences_of_twenty_words_mask.is_cuda)
pytorch_model = pytorch_model.cuda(device_id)
output_before_match = pytorch_model(five_sentences_of_twenty_words, five_sentences_of_twenty_words_mask)
assert output_before_match.shape == (5, some_random_d_model)
print(type(output_before_match))
print(output_before_match.is_cuda, output_before_match.get_device())
assert output_before_match.is_cuda
assert five_sentences_of_twenty_words.is_cuda
assert five_sentences_of_twenty_words_mask.is_cuda
我还在模块中使用了add_module方法： https://discuss.pytorch.org/t/when-to-use-add-module-function/10534
答案 2 :(得分：0)
要将“ CPU”张量转换为“ GPU”张量，只需执行以下操作：
<!DOCTYPE HTML>
<html>
<head>
<link rel="apple-touch-icon" sizes="180x180" href="images\free_horizontal_on_white_by_logaster.jpg">
<link rel="icon" type="image/jpg" sizes="32x32" href="images\free_horizontal_on_white_by_logaster.jpg">
<link rel="icon" type="image/jpg" sizes="16x16" href="images\free_horizontal_on_white_by_logaster.jpg">
<meta name="msapplication-TileColor" content="#da532c">
<meta name="theme-color" content="#ffffff">
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta http-equiv="X-UA-Compatible" content="ie=edge">
<link rel="stylesheet" href="resolve.css">
<title>Resolve - Real Women, Real Feedback</title>
</head>
<body>
<header>
<div class="container">
<div id="branding">
<a href="indexresolve.html"><img src="images/lasttry.png" alt="resolvelogo"></a>
</div>
<nav>
<li><a href="indexresolve.html">Home</a></li>
< <li><a href="howitworks.html">How It Works</a></li>
<li><a href="contact.html">Contact</a></li>
<li><a href="faq.html">FAQ</a></li>
<li><button id="login" class="button">Log In</button></li>
<div id="login-modal">
<div id="login-content">
<span class="close">×</span>
<img id="login-logo" src="images\free_horizontal_on_white_by_logaster.jpg">
<form>
<input class="login-input" type="text" placeholder="username">
<input class="login-input" type="password" placeholder="password">
<button>Log In</button>
</form>
<p>By clicking log in, you agree to our <a href="terms.html">Terms</a>, <a href="privacy.html">Privacy Policy</a>, and our <a href="cookie.html">Cookie Policy</a>.</p>
</div>
</div>
</nav>
</header>
<section>
<div class="container2">
<div>
<h1>Guys</h1>
<h2>fajfsda klfsdajfodisjflkd oisdjfklewjf oisdjfsakfj akfjfslkdja;fj sd;akfjdkfjsdakfj saifjsdakfs.</h2>
<button>Get Started</button>
</div>
<div>
<h1>Ladies</h1>
<h2>dklasdjfs kdsjdlk jfsalkjf las;fjdaa fdaksjdk skjfsidjf akldfjskl fjsdlkfjskdlfjsdifjdkf dkfjsdijf s </h2>
<button id="login">Get Started</button>
</div>
<div class="appad">
<h2>App Coming Soon!</h2>
</div>
</div>
<script src="resolve.js"></script>
</body>
</html>这会将张量带到默认的GPU设备。如果您有多个这样的GPU设备，那么您也可以像这样传递device_id：
cpuTensor = cpuTensor.cuda()