You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 7, 2023. It is now read-only.
When run_eval_autoregressive in hparams is set to True, Transformer in translation ende task always report the following error:
tensorflow.python.framework.errors_impl.InvalidArgumentError: The node 'cond_1/padded_accuracy/pad_with_zeros/pad_to_same_length/Shape_1/Switch' has inputs from different frames. The input 'Equal_1' is in frame 'while/while_context'. The input 'IteratorGetNext' is in frame ''.
The error is reproducible in various conditions with current version of tensor2tensor.
The text was updated successfully, but these errors were encountered:
When run_eval_autoregressive in hparams is set to True, Transformer in translation ende task always report the following error:
tensorflow.python.framework.errors_impl.InvalidArgumentError: The node 'cond_1/padded_accuracy/pad_with_zeros/pad_to_same_length/Shape_1/Switch' has inputs from different frames. The input 'Equal_1' is in frame 'while/while_context'. The input 'IteratorGetNext' is in frame ''.
The error is reproducible in various conditions with current version of tensor2tensor.
The text was updated successfully, but these errors were encountered: