mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-12-11 06:55:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix wenetspeech tokenizer

Browse Source
This commit is contained in:
pkufool 2023-05-16 18:37:21 +08:00
parent 949e49eec8
commit d7daa3b718
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
egs/wenetspeech/ASR/pruned_transducer_stateless5/decode.py
View File

@ -92,7 +92,7 @@ When training with the L subset, the streaming usage:
--causal-convolution 1 \ --causal-convolution 1 \
--decode-chunk-size 16 \ --decode-chunk-size 16 \
--left-context 64 --left-context 64
(4) modified beam search with RNNLM shallow fusion (4) modified beam search with RNNLM shallow fusion
./pruned_transducer_stateless5/decode.py \ ./pruned_transducer_stateless5/decode.py \
--epoch 35 \ --epoch 35 \
@ -853,7 +853,12 @@ def main():
if os.path.exists(params.context_file): if os.path.exists(params.context_file):
contexts = [] contexts = []
for line in open(params.context_file).readlines(): for line in open(params.context_file).readlines():
contexts.append(graph_compiler.texts_to_ids(line.strip())) context_list = graph_compiler.texts_to_ids(line.strip())
tmp = []
for context in context_list:
for x in context:
tmp.append(x)
contexts.append(tmp)
context_graph = ContextGraph(params.context_score) context_graph = ContextGraph(params.context_score)
context_graph.build(contexts) context_graph.build(contexts)
else: else: