麻烦贴下完整的报错和代码吧。

好的，我正在用这个代码参加一个比赛，请不要发到公开平台上，多谢啦 下图是报错代码： 复制错误信息如下： Exception has occurred: ValueError In transformed code: File "h:\Desktop\pp\evaluationNet.py", line 76, in forward x = F.relu(x) x = self.conv1(x) ~~~~~~~~~~~~~~~~~ <--- HERE x = x + x1 x = F.relu(x) File "C:\ProgramData\Anaconda3\envs\paddle_env\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call return self._dygraph_call_func(inputs, kwargs) File "C:\ProgramData\Anaconda3\envs\paddle_env\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(inputs, kwargs) File "C:\ProgramData\Anaconda3\envs\paddle_env\lib\site-packages\paddle\nn\layer\conv.py", line 347, in forward data_format=self._data_format) File "C:\ProgramData\Anaconda3\envs\paddle_env\lib\site-packages\paddle\nn\functional\conv.py", line 326, in conv1d format(x.shape)) ValueError: Input x should be 3D tensor, but received x with the shape of (-1, 32) File "C:\Users\congcong\AppData\Local\Temp\tmpy0tbyb73.py", line 13, in forward x = paddle.jit.dy2static.convert_call(self.conv1)(x) During handling of the above exception, another exception occurred: File "H:\Desktop\pp\evaluationNet.py", line 182, in <module> input_spec=[InputSpec(shape=[None, 15], dtype='float32')]) 完整代码如下： import paddle import paddle.nn.functional as F from paddle.io import Dataset from paddle.static import InputSpec #print(paddle.version) import numpy as np import pandas as pd from matplotlib import pyplot as plt from sklearn.model_selection import train_test_split import warnings warnings.filterwarnings("ignore") ev_path = 'H:/Desktop/pp/evaluation/Balance600.csv' ev = pd.read_csv(ev_path, index_col="FID") print(ev.head()) ''' fig, ax = plt.subplots() ev.plot(legend=False, ax=ax) plt.show() ''' data =  np.array(ev) print(data) datay = data[:,0]  #标签列 print(datay) data = (data - data.mean(axis=0))/data.std(axis=0)  #Zscore标准化 #data = (data-data.min(axis=0))/(data.max(axis=0)-data.min(axis=0)) data[:,0] = datay print(data) train, test = train_test_split(data, test_size = 0.2,random_state=0,shuffle=True,stratify=datay) print(train) print(test) # define a  dataset class MyDataset(Dataset):     def init(self, data):                 self.data = paddle.to_tensor(data, dtype='float32')     def getitem(self, idx):         data = self.data[idx,1:]         label = self.data[idx,0]         return data, label     def len(self):         return len(self.data) train_dataset = MyDataset(train) valid_dataset = MyDataset(test)     class MyNet(paddle.nn.Layer):     def init(self, num_classes=1):         super(MyNet, self).init()                 self.linear1 = paddle.nn.Linear(in_features=15, out_features=128)         self.linear2 = paddle.nn.Linear(in_features=128, out_features=32)         self.linear3 = paddle.nn.Linear(in_features=15, out_features=5)               self.conv1 = paddle.nn.Conv1D(in_channels=32, out_channels=15, kernel_size=1,data_format='NLC')         def forward(self, x):                 x1 = x         x = self.linear1(x)         x = F.relu(x)         x = F.dropout(x,0.016)                 x = self.linear2(x)         x = F.relu(x)                 x = self.conv1(x)         x = x + x1         x = F.relu(x)                 x = F.dropout(x,0.001)         x = self.linear3(x)               return x       # 参数设置 epoch_num = 100 batch_size = 138 learning_rate = 0.001 def train(model):     print('训练开始')     # 实例化模型     model = MyNet()     # 将模型转换为训练模式     model.train()     # 设置优化器，学习率，并且把模型参数给优化器     opt = paddle.optimizer.Adam(learning_rate=learning_rate,weight_decay=0.01,parameters=model.parameters())     # 设置损失函数     mse_loss = paddle.nn.MSELoss()     # 设置数据读取器     train_loader = paddle.io.DataLoader(train_dataset,                         batch_size=batch_size,                         shuffle=True)     valid_loader = paddle.io.DataLoader(valid_dataset, batch_size=batch_size)     val_acc_history = []     val_loss_history = []     for epoch in range(epoch_num):         for batch_id, data in enumerate(train_loader()):                         #x = data[0].t()  #矩阵转置             x = paddle.unsqueeze(data[0],-2)             #print(type(x))             #print(x.shape)             y = paddle.to_tensor(data[1],dtype='int64')             #print(y.shape)                         logits = model(x)             #print(logits.shape)             #print(y.shape)                         loss = F.cross_entropy(logits, y, soft_label=False, axis=-1)             if batch_id % 1000 == 0:                 print("epoch: {}, batch_id: {}, loss is: {}".format(epoch, batch_id, loss.numpy()))             loss.backward()             opt.step()             opt.clear_grad()         model.eval()         accuracies = []         losses = []           for batch_id, data in enumerate(valid_loader()):             x_data = paddle.unsqueeze(data[0],-2)             y_data = paddle.to_tensor(data[1],dtype='int64')                         logits = model(x_data)                         loss = F.cross_entropy(logits, y_data)             acc = paddle.metric.accuracy(logits, y_data)             #accf = paddle.metric.accuracy()             #acc = accf.compute(logits, y_data)             accuracies.append(acc.numpy())             losses.append(loss.numpy())               #print("predict:")             #print(logits)             #print("trueLabel:")             #print(y_data)                     avg_acc, avg_loss = np.mean(accuracies), np.mean(losses)         print("[validation] accuracy/loss: {}/{}".format(avg_acc, avg_loss))                 val_acc_history.append(avg_acc)         val_loss_history.append(avg_loss)         model.train()     plt.plot(val_acc_history, label = 'validation accuracy')     plt.xlabel('Epoch')     plt.ylabel('Accuracy')     plt.ylim([0.5, 0.8])     plt.legend(loc='lower right')     plt.show()         plt.plot(val_loss_history, label = 'validation loss')     plt.xlabel('Epoch')     plt.ylabel('Loss')     plt.ylim([0.5, 2])     plt.legend(loc='lower right')     plt.show() layer = MyNet() model = MyNet(num_classes=5) train(model) # save path = "H:/Desktop/pp/evaluation/evNet" paddle.jit.save(     layer=layer,     path=path,     input_spec=[InputSpec(shape=[None, 15], dtype='float32')]) 钟凤媛 ***@.  …

------------------ 原始邮件 ------------------ 发件人: "PaddlePaddle/Paddle" ***@***.***>; 发送时间: 2022年8月24日(星期三) 下午3:17 ***@***.***>; ***@***.******@***.***>; 主题: Re: [PaddlePaddle/Paddle] 动态图本地模型保存报错，Exception has occurred: ValueError In transformed code (Issue #45378) 麻烦贴下完整的报错和代码吧。 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

看下提示 ValueError: Input x should be 3D tensor, but received x with the shape of (-1, 32)

也麻烦格式化下代码和报错信息。

您好，辛苦您了，所报错的问题（ValueError: Input x should be 3D tensor, but received x with the shape of (-1, 32)）正是令人困惑的地方，这个程序已经训练测试成功，并没有报错，只是加了save部分才有报错信息。我把运行过程做了断点，显示这一句输入张量为三维，shape为[138,1,32], 您请看运行截图如下： 下一步也设上断点，还是三维张量： 之前那个格式可能因为粘贴到mail里面有丢失，我把代码文件放在附件里面，请查收。 钟凤媛 @***.  …

------------------ 原始邮件 ------------------ 发件人: "PaddlePaddle/Paddle" ***@***.***>; 发送时间: 2022年8月24日(星期三) 晚上7:21 ***@***.***>; ***@***.******@***.***>; 主题: Re: [PaddlePaddle/Paddle] 动态图本地模型保存报错，Exception has occurred: ValueError In transformed code (Issue #45378) 看下提示 ValueError: Input x should be 3D tensor, but received x with the shape of (-1, 32) 也麻烦格式化下代码和报错信息。 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

check下inputspec中给的shape是否正确。应该就是shape的问题。

非常感谢，正是shape设置的问题，仔细查看了函数的说明，发现了shape参数的详细要求，问题解决了。给你100个赞! 钟凤媛 @***.  …

------------------ 原始邮件 ------------------ 发件人: "PaddlePaddle/Paddle" ***@***.***>; 发送时间: 2022年8月25日(星期四) 上午10:32 ***@***.***>; ***@***.******@***.***>; 主题: Re: [PaddlePaddle/Paddle] 动态图本地模型保存报错，Exception has occurred: ValueError In transformed code (Issue #45378) check下inputspec中给的shape是否正确。应该就是shape的问题。 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>