I am getting this error when running fit():

Node: 'gradient_tape/sequential/dense_1/MatMul/MatMul'
Matrix size-incompatible: In[0]: [32,4], In[1]: [16,1]
         [[{{node gradient_tape/sequential/dense_1/MatMul/MatMul}}]] [Op:__inference_train_function_707]

where should I look for the error ?
Is it in the model definition ?
Is it related to the training dataset ?
Is it related to the validation dataset ?
Is it the loss function ?

I need an idea of where to start looking

here is the entire error if it helps

Traceback (most recent call last):
  File "E:Projectstftest4.py", line 169, in <module>
    model.fit(traingen,
  File "C:Pythonlibsite-packageskerasutilstraceback_utils.py", line 67, in error_handler
    raise e.with_traceback(filtered_tb) from None
  File "C:Pythonlibsite-packagestensorflowpythoneagerexecute.py", line 54, 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 'gradient_tape/sequential/dense_1/MatMul/MatMul' defined at (most recent call last):
    File "E:Projectstftest4.py", line 169, in <module>
      model.fit(traingen,
    File "C:Pythonlibsite-packageskerasutilstraceback_utils.py", line 64, in error_handler
      return fn(*args, **kwargs)
    File "C:Pythonlibsite-packageskerasenginetraining.py", line 1409, in fit
      tmp_logs = self.train_function(iterator)
    File "C:Pythonlibsite-packageskerasenginetraining.py", line 1051, in train_function
      return step_function(self, iterator)
    File "C:Pythonlibsite-packageskerasenginetraining.py", line 1040, in step_function
      outputs = model.distribute_strategy.run(run_step, args=(data,))
    File "C:Pythonlibsite-packageskerasenginetraining.py", line 1030, in run_step
      outputs = model.train_step(data)
    File "C:Pythonlibsite-packageskerasenginetraining.py", line 893, in train_step
      self.optimizer.minimize(loss, self.trainable_variables, tape=tape)
    File "C:Pythonlibsite-packageskerasoptimizersoptimizer_v2optimizer_v2.py", line 537, in minimize
      grads_and_vars = self._compute_gradients(
    File "C:Pythonlibsite-packageskerasoptimizersoptimizer_v2optimizer_v2.py", line 590, in _compute_gradients
      grads_and_vars = self._get_gradients(tape, loss, var_list, grad_loss)
    File "C:Pythonlibsite-packageskerasoptimizersoptimizer_v2optimizer_v2.py", line 471, in _get_gradients
      grads = tape.gradient(loss, var_list, grad_loss)
Node: 'gradient_tape/sequential/dense_1/MatMul/MatMul'
Matrix size-incompatible: In[0]: [32,4], In[1]: [16,1]
         [[{{node gradient_tape/sequential/dense_1/MatMul/MatMul}}]] [Op:__inference_train_function_707]