最后，介绍一下模型的训练、验证以及测试流程。
目录
1. 训练
def train(config, model, train_iter, dev_iter, test_iter):
start_time = time.time()
model.train()#训练模式
param_optimizer = list(model.named_parameters())
#下列参数 不进行正则化（权重衰减）
no_decay = ['bias', 'LayerNorm.bias', 'LayerNorm.weight']
optimizer_grouped_parameters = [
{'params': [p for n, p in param_optimizer if not any(nd in n for nd in no_decay)], 'weight_decay': 0.01},
{'params': [p for n, p in param_optimizer if any(nd in n for nd in no_decay)], 'weight_decay': 0.0}]
# optimizer = torch.optim.Adam(model.parameters(), lr=config.learning_rate)
#优化器
optimizer = BertAdam(optimizer_grouped_parameters,
lr=config.learning_rate,
warmup=0.05,
t_total=len(train_iter) * config.num_epochs)
total_batch = 0 # 记录进行到多少batch
dev_best_loss = float('inf')#跟踪验证集最小的loss 或 最大f1-score或准确率
last_improve = 0 # 记录上次验证集loss下降的batch数
flag = False # 记录是否很久没有效果提升
model.train() #训练模式
for epoch in range(config.num_epochs):
print('Epoch [{}/{}]'.format(epoch + 1, config.num_epochs))
for i, (trains, labels) in enumerate(train_iter):
outputs = model(trains)#前向传播 获取输出
model.zero_grad()#清空梯度
loss = F.cross_entropy(outputs, labels)#计算交叉熵损失（内部包含softmax log等操作） 可以用nn里面的函数 也可以用F中的函数 labels为整数索引 内部会自动转换为one-hot
loss.backward()#计算梯度
optimizer.step()#更新参数
if total_batch % 100 == 0:
#每100个batch 计算一下在验证集上的指标 或者像之前项目中那样 每一个epoch在验证集上计算相关指标
true = labels.data.cpu()
predic = torch.max(outputs.data, 1)[1].cpu()
train_acc = metrics.accuracy_score(true, predic)#当前batch上训练集的准确率 因为是类别均衡数据集，所以可以直接用准确率作评估指标
dev_acc, dev_loss = evaluate(config, model, dev_iter)#计算此时模型在验证集上的损失和准确率
if dev_loss < dev_best_loss:
dev_best_loss = dev_loss#更新验证集最小损失
torch.save(model.state_dict(), config.save_path)#保存在验证集上损失最小的参数
improve = '*'#效果提升标志
last_improve = total_batch
else:
improve = ''
time_dif = get_time_dif(start_time)
msg = 'Iter: {0:>6}, Train Loss: {1:>5.2}, Train Acc: {2:>6.2%}, Val Loss: {3:>5.2}, Val Acc: {4:>6.2%}, Time: {5} {6}'
print(msg.format(total_batch, loss.item(), train_acc, dev_loss, dev_acc, time_dif, improve))
model.train()#回到训练模式
total_batch += 1
if total_batch - last_improve > config.require_improvement:
# 如果长期没有提高 就提前终止
# 验证集loss超过1000batch没下降，结束训练
print("No optimization for a long time, auto-stopping...")
flag = True
break
if flag:
break
test(config, model, test_iter)#模型训练结束后 进行测试2. 验证
def evaluate(config, model, data_iter, test=False):
model.eval()#测试模式
loss_total = 0
predict_all = np.array([], dtype=int)#存储验证集所有batch的预测结果
labels_all = np.array([], dtype=int)#存储验证集所有batch的真实标签
with torch.no_grad():
for texts, labels in data_iter:
outputs = model(texts)
loss = F.cross_entropy(outputs, labels)
loss_total += loss
labels = labels.data.cpu().numpy()
predic = torch.max(outputs.data, 1)[1].cpu().numpy()
labels_all = np.append(labels_all, labels)
predict_all = np.append(predict_all, predic)
acc = metrics.accuracy_score(labels_all, predict_all)#计算验证集准确率
if test:#如果是测试集的话 计算一下分类报告
report = metrics.classification_report(labels_all, predict_all, target_names=config.class_list, digits=4)
confusion = metrics.confusion_matrix(labels_all, predict_all)#计算混淆矩阵
return acc, loss_total / len(data_iter), report, confusion
return acc, loss_total / len(data_iter)#返回准确率和每个batch的平均损失3. 测试
def test(config, model, test_iter):
# test
model.load_state_dict(torch.load(config.save_path))#加载使验证集损失最小的参数
model.eval()#测试模式
start_time = time.time()
test_acc, test_loss, test_report, test_confusion = evaluate(config, model, test_iter, test=True)#计算测试集准确率，每个batch的平均损失 分类报告、混淆矩阵
msg = 'Test Loss: {0:>5.2}, Test Acc: {1:>6.2%}'
print(msg.format(test_loss, test_acc))
print("Precision, Recall and F1-Score...")
print(test_report)
print("Confusion Matrix...")
print(test_confusion)
time_dif = get_time_dif(start_time)
print("Time usage:", time_dif)