|
From: E <oth...@ao...> - 2013-10-17 03:21:48
|
Hi!
When I run online decoder I see below output on screen
online-audio-server-decode-faster --verbose=5 --rt-min=0.7 --rt-max=0.9 --max-active=10000 --beam=72.0 --acoustic-scale=0.0769 ./final.mdl graph_tgpr/HCLG.fst graph_tgpr/words.txt 1:2:3:4:5:6:7:8:9:10:11:12:13:14:15 5010 graph_tgpr/phones/word_boundary.int final.mat
TcpServer: Listening on port: 5010
Reading LDA matrix: final.mat...
Reading acoustic model: ./final.mdl...
Reading word list: graph_tgpr/words.txt...
Reading word boundary file: graph_tgpr/phones/word_boundary.int...
Reading FST: graph_tgpr/HCLG.fst...
Waiting for client...
Then it stays like that for a long time without recognizing. Could this be because the HCLG.fst file is too large (585M)?
I checked mic by recording in Audacity. Recording works.
How to debug this issue?
Thanks!
|