本文整理匯總了Python中net.Net.eval方法的典型用法代碼示例。如果您正苦於以下問題：Python Net.eval方法的具體用法？Python Net.eval怎麽用？Python Net.eval使用的例子？那麽, 這裏精選的方法代碼示例或許可以為您提供幫助。您也可以進一步了解該方法所在類net.Net的用法示例。
在下文中一共展示了Net.eval方法的1個代碼示例，這些例子默認根據受歡迎程度排序。您可以為喜歡或者感覺有用的代碼點讚，您的評價將有助於係統推薦出更棒的Python代碼示例。
示例1: Solver
# 需要導入模塊: from net import Net [as 別名]
# 或者: from net.Net import eval [as 別名]
class Solver():
def __init__(self, args):
# prepare shakespeare dataset
train_iter, data_info = load_shakespeare(args.batch_size, args.bptt_len)
self.vocab_size = data_info["vocab_size"]
self.TEXT = data_info["TEXT"]
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
self.net = Net(self.vocab_size, args.embed_dim,
args.hidden_dim, args.num_layers).to(self.device)
self.loss_fn = torch.nn.CrossEntropyLoss(ignore_index=1) # <pad>: 1
self.optim = torch.optim.Adam(self.net.parameters(), args.lr)
self.args = args
self.train_iter = train_iter
if not os.path.exists(args.ckpt_dir):
os.makedirs(args.ckpt_dir)
def fit(self):
args = self.args
for epoch in range(args.max_epochs):
self.net.train()
for step, inputs in enumerate(self.train_iter):
X = inputs.text.to(self.device)
y = inputs.target.to(self.device)
out, _ = self.net(X)
loss = self.loss_fn(out, y.view(-1))
self.optim.zero_grad()
loss.backward()
self.optim.step()
if (epoch+1) % args.print_every == 0:
text = self.sample(args.sample_length, args.sample_prime)
print("Epoch [{}/{}] loss: {:.3f}"
.format(epoch+1, args.max_epochs, loss.item()/args.bptt_len))
print(text, "\n")
self.save(args.ckpt_dir, args.ckpt_name, epoch+1)
def sample(self, length, prime="The"):
args = self.args
self.net.eval()
samples = list(prime)
# convert prime string to torch.LongTensor type
prime = self.TEXT.process(prime, device=self.device, train=False)
# sample character indices
indices = self.net.sample(prime, length)
# convert char indices to string type
for index in indices:
out = self.TEXT.vocab.itos[index.item()]
samples.append(out.replace("<eos>", "\n"))
self.TEXT.sequential = True
return "".join(samples)
def save(self, ckpt_dir, ckpt_name, global_step):
save_path = os.path.join(
ckpt_dir, "{}_{}.pth".format(ckpt_name, global_step))
torch.save(self.net.state_dict(), save_path)