# include common settings for fullLP systems.
. conf/common.fullLP || exit 1;

#speech corpora files location
train_data_dir=/export/babel/data/101-cantonese/release-current/conversational/training
train_data_list=/export/babel/data/splits/Cantonese_Babel101/train.FullLP.list
train_nj=32

#RADICAL DEV data files
dev2h_data_dir=/export/babel/data/101-cantonese/release-current/conversational/dev
dev2h_data_list=/export/babel/data/splits/Cantonese_Babel101/dev.3hr.list
dev2h_data_cmudb=/export/babel/data/splits/Cantonese_Babel101/uem/db-v8-utt.dat
dev2h_stm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev/IARPA-babel101b-v0.4c_conv-dev.stm
dev2h_ecf_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev.ecf.xml
dev2h_rttm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev/IARPA-babel101b-v0.4c_conv-dev.mitllfa2.rttm
dev2h_kwlist_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev.kwlist.xml
dev2h_subset_ecf=true
dev2h_nj=20

#Official DEV data files
dev10h_data_dir=/export/babel/data/101-cantonese/release-current/conversational/dev
dev10h_data_list=/export/babel/data/splits/Cantonese_Babel101/dev.list
dev10h_data_cmudb=/export/babel/data/splits/Cantonese_Babel101/uem/db-v8-utt.dat
dev10h_stm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev/IARPA-babel101b-v0.4c_conv-dev.stm
dev10h_ecf_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev.ecf.xml
dev10h_rttm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev/IARPA-babel101b-v0.4c_conv-dev.mitllfa2.rttm
dev10h_kwlist_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-dev.kwlist.xml
dev10h_nj=32


#Official EVAL period evaluation data files
eval_data_dir=/export/babel/data/101-cantonese/release-current/conversational/eval
eval_data_list=/export/babel/data/splits/Cantonese_Babel101/eval.babel101b-v0.4c.list
eval_data_cmudb=/export/babel/data/splits/Cantonese_Babel101/uem/db-v8-utt.dat
eval_ecf_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-eval.ecf.xml
eval_kwlist_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-eval.kwlist.xml
eval_nj=64

#Official (POST-)EVAL evaluation data portion
evalpart1_data_dir=/export/babel/data/101-cantonese/release-current/conversational/eval
evalpart1_data_list=/export/babel/data/splits/Cantonese_Babel101/evalpart1.babel101b-v0.4c.list
evalpart1_data_cmudb=/export/babel/data/splits/Cantonese_Babel101/uem/db-v8-utt.dat
evalpart1_stm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-evalpart1/IARPA-babel101b-v0.4c_conv-evalpart1.stm
evalpart1_ecf_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-evalpart1.ecf.xml
evalpart1_rttm_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-evalpart1/IARPA-babel101b-v0.4c_conv-evalpart1.mitllfa3.rttm
evalpart1_kwlist_file=/export/babel/data/scoring/IndusDB/IARPA-babel101b-v0.4c_conv-evalpart1/IARPA-babel101b-v0.4c_conv-evalpart1.annot.kwlist.xml
evalpart1_nj=21

# Acoustic model parameters
numLeavesTri1=1000
numGaussTri1=10000
numLeavesTri2=1000
numGaussTri2=20000
numLeavesTri3=6000
numGaussTri3=75000
numLeavesMLLT=6000
numGaussMLLT=75000
numLeavesSAT=6000
numGaussSAT=75000
numGaussUBM=800
numLeavesSGMM=10000
numGaussSGMM=80000


# Lexicon and Language Model parameters
oovSymbol="<unk>"
lexiconFlags="--romanized --oov <unk>"

# Scoring protocols (dummy GLM file to appease the scoring script)
#glmFile=/export/babel/data/splits/Cantonese_Babel101/cantonese.glm
lexicon_file=/export/babel/data/101-cantonese/release-current/conversational/reference_materials/lexicon.txt
cer=1

max_index_states=150000
word_ins_penalty=0.5

#keyword search settings
duptime=0.5
case_insensitive=true