I exported a trained transformer
model for the enes_wmt32k
problem with the following
t2t-exporter
--model=transformer
--hparams_set=transformer_tiny
--problem=translate_enes_wmt32k
--data_dir=<PATH_TO_DATA_DIR>
--output_dir=<PATH_TO_TRAIN_DIR>
I then started the tensorflow_model_server as such
I ensured default_serving
signature is present in the exported model saved_model.pbtxt
as such
However I get an input parsing error when pinging the REST API as such:
Do you know why is this happening?
The details of my environment are here:
Thank you, Martin,It is still a bad request (see picture)
If so, do you have any idea which inputs does it expect? The result from the encode method, maybe? - explained in hereThank you,