You need to sign in or sign up before continuing.
Commit 4a749e2d authored by Zhaoheng Ni's avatar Zhaoheng Ni Committed by Facebook GitHub Bot
Browse files

Fix arguments in CTC decoding script (#2315)

Summary:
Some arguments in `ArgumentParser` are not used in the `lexicon_decoder`. Fix them to use the ones in the parser.

Pull Request resolved: https://github.com/pytorch/audio/pull/2315

Reviewed By: carolineechen

Differential Revision: D35357678

Pulled By: nateanl

fbshipit-source-id: 4e70418cf03708b82bc158cafd9999a80ad08f92
parent 87f0d198
...@@ -35,14 +35,14 @@ def run_inference(args): ...@@ -35,14 +35,14 @@ def run_inference(args):
lexicon=lexicon_file, lexicon=lexicon_file,
tokens=tokens, tokens=tokens,
lm=kenlm_file, lm=kenlm_file,
nbest=1, nbest=args.nbest,
beam_size=1500, beam_size=args.beam_size,
beam_size_token=None, beam_size_token=args.beam_size_token,
beam_threshold=50, beam_threshold=args.beam_threshold,
lm_weight=args.lm_weight, lm_weight=args.lm_weight,
word_score=args.word_score, word_score=args.word_score,
unk_score=float("-inf"), unk_score=args.unk_score,
sil_score=0, sil_score=args.sil_score,
log_add=False, log_add=False,
) )
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment