训练模型时报错: Incompatible shapes: [14] vs. [14,50]

[DragonwolfAside]

环境：
Python 3.10.10
Tensorflow 2.12.0

日志：

Tue May 23 12:23:08 2023 正在处理训练数据...
Tue May 23 12:23:08 2023 循环轮数:5 batch size:20
Epoch 1/5
Traceback (most recent call last):
  File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\main.py", line 35, in <module>
    main()
  File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\main.py", line 22, in main
    seq.train_model(size, epoch)
  File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\seq2seq.py", line 79, in train_model
    model.fit([encoder_input_data, decoder_input_data], decoder_target_data,
  File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\utils\traceback_utils.py", line 70, in error_handler
    raise e.with_traceback(filtered_tb) from None
  File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\tensorflow\python\eager\execute.py", line 52, in quick_execute
    tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,
tensorflow.python.framework.errors_impl.InvalidArgumentError: Graph execution error:

Detected at node 'Equal' defined at (most recent call last):
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\main.py", line 35, in <module>
      main()
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\main.py", line 22, in main
      seq.train_model(size, epoch)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\seq2seq.py", line 79, in train_model
      model.fit([encoder_input_data, decoder_input_data], decoder_target_data,
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\utils\traceback_utils.py", line 65, in error_handler
      return fn(*args, **kwargs)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1685, in fit
      tmp_logs = self.train_function(iterator)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1284, in train_function
      return step_function(self, iterator)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1268, in step_function
      outputs = model.distribute_strategy.run(run_step, args=(data,))
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1249, in run_step
      outputs = model.train_step(data)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1055, in train_step
      return self.compute_metrics(x, y, y_pred, sample_weight)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\training.py", line 1149, in compute_metrics
      self.compiled_metrics.update_state(y, y_pred, sample_weight)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\engine\compile_utils.py", line 605, in update_state
      metric_obj.update_state(y_t, y_p, sample_weight=mask)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\utils\metrics_utils.py", line 77, in decorated
      update_op = update_state_fn(*args, **kwargs)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\metrics\base_metric.py", line 140, in update_state_fn
      return ag_update_state(*args, **kwargs)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\metrics\base_metric.py", line 691, in update_state
      matches = ag_fn(y_true, y_pred, **self._fn_kwargs)
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\metrics\accuracy_metrics.py", line 426, in categorical_accuracy
      return metrics_utils.sparse_categorical_matches(
    File "C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot\venv\lib\site-packages\keras\utils\metrics_utils.py", line 971, in sparse_categorical_matches
      matches = tf.cast(tf.equal(y_true, y_pred), backend.floatx())
Node: 'Equal'
Incompatible shapes: [14] vs. [14,50]
         [[{{node Equal}}]] [Op:__inference_train_function_19230]

(venv) C:\Users\Administrator\Desktop\NLP\OSSAS Chatbot>

[3588044667HZ]

我的也这样

[RepentStar]

代码执行错误提示“InvalidArgumentError: Graph execution error”通常是因为TensorFlow计算图中的某个操作失败了。具体到这段代码，错误是在train_model函数中触发的，而错误的根本原因是“Detected at node Equal defined at”和“Incompatible shapes: [2] vs. [2,15]”。
这表明在计算图中有两个形状不兼容的Tensor在进行比较操作（Equal节点）。一个形状是[2]，另一个是[2,15]，显然这两个形状不能进行比较。
要解决这个问题，你需要检查train_model函数中所有进行比较的操作，确保比较的Tensor具有相同的形状。特别是，需要检查在模型编译阶段定义的损失函数和评价指标，以及在训练过程中使用的任何自定义函数或层。
你可以通过以下步骤来调试这个问题：
确认在模型编译阶段，损失函数和评价指标使用的Tensor形状是否一致。
如果使用自定义层或函数，检查这些层或函数的输入和输出Tensor的形状是否匹配。
在训练循环中，检查传入模型的数据（输入）和模型的预测结果（输出）的形状是否一致。
解决上述问题后，错误应该会得到解决。如果问题依然存在，可能需要进一步检查模型的架构，以确保在训练过程中所有Tensor的形状都是预期的。
以上内容由AI回答。
这个项目几年来并没有任何变化，两年前我训练也是正常的，不知道为什么现在就不行了，报错类型和这个issue一样，训练数据一模一样，不知道哪里出了问题
注：我两年前用的是.exe进行的训练，现在出现这个问题的好像都是用源代码训练的