with tpu_strategy.scope():
for x in dataset1.as_numpy_iterator():
g1=siamese_model.train_step(x)
break
it is giving error : SystemError: <built-in method TFE_Py_TapeVariableAccessed of PyCapsule object at 0x7f2b701858a0> returned a result with an error set
link to notebook:tpu_fashion_try1 | Kaggle
please if anyone can help resolve the error