ictnlp / ddrs-nat Goto Github PK

Thank you for your work!
When running the shell in **Inference ** Step 1: Decode the test set with argmax decoding, I faced the error as follows:
Traceback (most recent call last):
File "generate.py", line 383, in
cli_main()
File "generate.py", line 379, in cli_main
main(args)
File "generate.py", line 41, in main
return main(args, sys.stdout)
File "generate.py", line 191, in main
hypos = task.inference_step(
File "/data1/workspace/DDRS-NAT/fairseq/tasks/fairseq_task.py", line 433, in inference_step
return generator.generate(
File "/root/anaconda3/envs/nmt/lib/python3.8/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "/data1/workspace/DDRS-NAT/fairseq/iterative_refinement_generator.py", line 229, in generate
log_probs, decoder_out, , = model.forward_decoder(
ValueError: too many values to unpack (expected 2)
and
I checked this line, found that
log_probs, decoder_out = model.forward_decoder( prev_decoder_out, encoder_out, **decoder_options )
the log_probs is missing.
How can I get the right results?