Question

这是一个运行基于字符的语言生成的RNN模型：

class RNN(nn.Module):
    def __init__(self, input_size, hidden_size, output_size, n_layers):
        super(RNN, self).__init__()
        self.input_size = input_size
        self.hidden_size = hidden_size
        self.output_size = output_size
        self.n_layers = n_layers 

        self.encoder = nn.Embedding(input_size, hidden_size)
        self.GRU = nn.GRU(hidden_size, hidden_size, n_layers, batch_first=True) 
        self.decoder = nn.Linear(hidden_size, output_size)


    def forward(self, input, batch_size):
        self.init_hidden(batch_size)
        input = self.encoder(input)
        output, self.hidden = self.GRU(input, self.hidden) 
        output = self.decoder(output.view(batch_size, self.hidden_size)) 
        return output

    def init_hidden(self, batch_size):
        self.hidden = Variable(torch.randn(self.n_layers, batch_size, self.hidden_size).cuda())

我使用DataParallel实例化模型，在我的4个GPU中分割输入批量：

net = torch.nn.DataParallel(RNN(n_chars, hidden_size, n_chars, n_layers)).cuda()

这里是full code。

不幸的是，DataParallel要求输入将batch_size作为第一维，但GRU函数需要隐藏张量将batch_size作为第二维：

output, self.hidden = self.GRU(input, self.hidden)

代码将引发以下错误（请注意打印输出显示编码器在4个GPU上正确执行）：

...
forward function: encoding input of shape: (16L, 1L)
forward function: encoding input of shape: (16L, 1L)
forward function: encoding input of shape: (16L,
forward function: encoding input of shape:

forward function: GRU processing input of shape:
1L)
 ( (16L, 16L1L, 1L), 100L)
forward function: GRU processing input of shape:
 (16L, 1L,
forward function: GRU processing input of shape:100L)
 (16L
forward function: GRU processing input of shape:, 1L, 100L) (
16L, 1L, 100L)

Traceback (most recent call last):
  File "gru2.py", line 166, in <module>
    output = net(c, batch_size)
  File "/root/miniconda2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 206, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/miniconda2/lib/python2.7/site-packages/torch/nn/parallel/data_parallel.py", line 61, in forward
    outputs = self.parallel_apply(replicas, inputs, kwargs)
  File "/root/miniconda2/lib/python2.7/site-packages/torch/nn/parallel/data_parallel.py", line 71, in parallel_apply
    return parallel_apply(replicas, inputs, kwargs)
  File "/root/miniconda2/lib/python2.7/site-packages/torch/nn/parallel/parallel_apply.py", line 45, in parallel_apply
    raise output
RuntimeError: Expected hidden size (2, 16L, 100), got (2L, 64L, 100L)

此处模型有2层，batch_size = 64，hidden_size = 100.

如何在前向功能中并行化GRU操作？

Answer 1

您只需设置参数dim = 1，例如

net = torch.nn.DataParallel(RNN(n_chars, hidden_size, n_chars, n_layers), dim=1).cuda()

Answer 2

PyTorch 1.5已完全解决了RNN培训和DataParallel的问题。似乎已经无缝地完成了。无需再进行gerrymandering。我今天在一个涉及语音mfcc的双向GRUS的项目中证实了这一点。


class PEncoder(nn.Module):
    def __init__(self, args, encoder):
        super(PEncoder, self).__init__()
        self.gpu_ids = args.gpu_ids
        self.model = encoder

    def forward(self, input):
        if len(self.gpu_ids) > 1:
            return nn.parallel.data_parallel(self.model, (input), self.gpu_ids)
        else:
            return self.model(input)

就是这么简单。这确实将您的模型包装到另一个模型中，并有效地产生了稍微不同的计算图。因此，如果您拥有较早训练的模型，则可能必须以特殊方式加载它们并为此并行包装创建一些设置器。尝试一下，您会看到的。（我尚未确认这方面。）

如何在Pytorch中将RNN函数与DataParallel

2 个答案: