代码块展示

代码块展示这里是各种代码块的展示示例：

import torch
dict=[‘e’,’h’,’l’,’o’]
x_data=[1,0,2,2,3]
x_data=torch.LongTensor([x_data])
y_data=torch.LongTensor([3,1,2,3,2])

class rnnmmodel(torch.nn.Module):
def init(self,dictionary_size,num_class):
super(rnnmmodel,self).init()
self.hidden_size=8
self.bedding_size=10
self.dictionary_size=dictionary_size
self.num_class=num_class
self.embedding=torch.nn.Embedding(self.dictionary_size,self.bedding_size)
self.rnn=torch.nn.RNN(input_size=self.bedding_size, hidden_size=self.hidden_size, num_layers=1, batch_first=True)
self.Linear=torch.nn.Linear(self.hidden_size, self.num_class)

def forward(self,x):
    h0=torch.zeros(1,x.size(0),self.hidden_size)
    x=self.embedding(x)
    x,_=self.rnn(x,h0)
    x=x.view(-1,self.hidden_size)
    x=self.Linear(x)
    return x

if name==’main‘:
model=rnnmmodel(4,4)
criterion=torch.nn.CrossEntropyLoss()
optimizer=torch.optim.Adam(model.parameters(),lr=0.01)
x_data=x_data.view(-1,5)
y_data=y_data.view(-1)
for epoch in range(15):
y_hat=model(x_data)
_,index=torch.max(y_hat,1)
index=index.data.numpy()
loss=criterion(y_hat,y_data)
print(‘epoch:’,epoch,’loss:’,loss.item(),’guess:’,’’.join([dict[x] for x in index]))
optimizer.zero_grad()
loss.backward()
optimizer.step()