One more issue from my end, that the test function seems to be running infinitely on the train model despite using “itertools.cycle”. @Mubsi
Its not supposed to happen because the n_steps = 1 in the train_model(model, data_generator, lines, eval_lines, batch_size=32, max_length=64, n_steps=1, output_dir=‘model/’). I dont know whats happening with your case though.
