Inputs to eager execution function cannot be Keras symbolic tensors

[neverfox]

I get the following error when trying to fit the model (Tensorflow 2.3.1):

Error in py_call_impl(callable, dots$args, dots$keywords) : 
  _SymbolicException: Inputs to eager execution function cannot be Keras symbolic tensors, but found [<tf.Tensor 'repeat_vector/Tile:0' shape=(None, 9, 49) dtype=float32>] 
22.
stop(structure(list(message = "_SymbolicException: Inputs to eager execution function cannot be Keras symbolic tensors, but found [<tf.Tensor 'repeat_vector/Tile:0' shape=(None, 9, 49) dtype=float32>]", 
    call = py_call_impl(callable, dots$args, dots$keywords), 
    cppstack = structure(list(file = "", line = -1L, stack = c("1   reticulate.so                       0x000000011f3c1524 _ZN4Rcpp9exceptionC2EPKcb + 196", 
    "2   reticulate.so                       0x000000011f3c91f5 _ZN4Rcpp4stopERKNSt3__112basic_stringIcNS0_11char_traitsIcEENS0_9allocatorIcEEEE + 53",  ... 
21.
quick_execute at execute.py#74
20.
call at function.py#550
19.
_call_flat at function.py#1924
18.
_filtered_call at function.py#1848
17.
__call__ at function.py#2829
16.
_call at def_function.py#807
15.
__call__ at def_function.py#780
14.
fit at training.py#1098
13.
_method_wrapper at training.py#108
12.
(structure(function (...) 
{
    dots <- py_resolve_dots(list(...))
    result <- py_call_impl(callable, dots$args, dots$keywords) ... 
11.
do.call(object$fit, args) 
10.
fit.keras.engine.training.Model(., x = full_training_data_keras$x, 
    y = full_training_data_keras$y, validation_data = unname(validation_data_keras), 
    batch_size = 2250, epochs = 1000, callbacks = list(cb), verbose = 0) 
9.
fit(., x = full_training_data_keras$x, y = full_training_data_keras$y, 
    validation_data = unname(validation_data_keras), batch_size = 2250, 
    epochs = 1000, callbacks = list(cb), verbose = 0) 
8.
function_list[[k]](value) 
7.
withVisible(function_list[[k]](value)) 
6.
freduce(value, `_function_list`) 
5.
`_fseq`(`_lhs`) 
4.
eval(quote(`_fseq`(`_lhs`)), env, env) 
3.
eval(quote(`_fseq`(`_lhs`)), env, env) 
2.
withVisible(eval(quote(`_fseq`(`_lhs`)), env, env)) 
1.
model %>% fit(x = full_training_data_keras$x, y = full_training_data_keras$y, 
    validation_data = unname(validation_data_keras), batch_size = 2250, 
    epochs = 1000, callbacks = list(cb), verbose = 0)

[AnnaDai1001]

Hello, I am having the same error with TensorFlow2.1.0. I am wondering whether you have fixed this issue. Thanks a lot!