├── recoMic ├── perf.sh ├── results_piBplus │ ├── results_PiBplus.tgz │ ├── results_14.txt │ ├── results_18.txt │ ├── results_20.txt │ ├── results_62.txt │ ├── results_47.txt │ ├── psphinx_10LM_PiBplus.log │ └── psphinx_14.log ├── results_pi3 │ ├── result_63.txt │ ├── result_14.txt │ ├── results_37.log │ ├── results_76.log │ ├── results_92.log │ ├── LMsmall_Pi3_10.txt │ ├── result_122.txt │ ├── results_66.log │ ├── mymain.log │ ├── results.log │ └── pi3_results.log ├── Pi3_Notes.txt ├── PiBplus_Notes.txt ├── perf.awk ├── pocket_sphinx_listen.py ├── recoMicNotes.txt ├── mymain.py ├── justreco.py ├── corpus.txt ├── dictionary.dic ├── grammar.jsgf ├── pocket_sphinx_listener.py ├── pocket_sphinx_listener.gram.py ├── pocket_sphinx_listener.lm.py └── language_model.lm ├── recoFile ├── perf.sh ├── test16k.wav ├── results │ ├── result_Pi3_file.log │ ├── result_PiBplus_file.log │ ├── reco_Pi3_file.log │ └── reco_PiBplus_file.log ├── input_file.txt ├── perf.awk └── recoFileNotes.txt ├── README.md └── Sphinx On Pi3.txt /recoMic/perf.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | awk -f perf.awk psphinx.log 3 | -------------------------------------------------------------------------------- /recoFile/perf.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | awk -f perf.awk psphinx.log 3 | -------------------------------------------------------------------------------- /recoFile/test16k.wav: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/slowrunner/Pi3RoadTest/HEAD/recoFile/test16k.wav -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_PiBplus.tgz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/slowrunner/Pi3RoadTest/HEAD/recoMic/results_piBplus/results_PiBplus.tgz -------------------------------------------------------------------------------- /recoMic/results_pi3/result_63.txt: -------------------------------------------------------------------------------- 1 | Utterances=63 2 | CpuTime=75.7 seconds 3 | CPU xRealTime=0.52 or 52% of one core 4 | Actual Speech=145.577 seconds 5 | Utterances=189.77 seconds total 6 | 77% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/result_14.txt: -------------------------------------------------------------------------------- 1 | Utterances=14 2 | CpuTime=56.42 seconds 3 | CPU xRealTime=0.926 or 92.6% of one core 4 | Actual Speech=60.9287 seconds 5 | Utterances=92.89 seconds total 6 | 66% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/results_37.log: -------------------------------------------------------------------------------- 1 | Utterances=37 2 | CpuTime=68.59 seconds 3 | CPU xRealTime=0.81 or 81% of one core 4 | Actual Speech=84.679 seconds 5 | Utterances=128.35 seconds total 6 | 66% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/results_76.log: -------------------------------------------------------------------------------- 1 | Utterances=76 2 | CpuTime=90.94 seconds 3 | CPU xRealTime=0.778 or 77.8% of one core 4 | Actual Speech=116.889 seconds 5 | Utterances=197.55 seconds total 6 | 59% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/results_92.log: -------------------------------------------------------------------------------- 1 | Utterances=92 2 | CpuTime=165.21 seconds 3 | CPU xRealTime=0.827 or 82.7% of one core 4 | Actual Speech=199.77 seconds 5 | Utterances=343.88 seconds total 6 | 58% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoFile/results/result_Pi3_file.log: -------------------------------------------------------------------------------- 1 | Utterances=15 2 | CpuTime=90.75 seconds 3 | CPU xRealTime=2.168 or 216.8% of one core 4 | Actual Speech=41.8589 seconds 5 | Utterances=90.95 seconds total 6 | 46% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/LMsmall_Pi3_10.txt: -------------------------------------------------------------------------------- 1 | Utterances=11 2 | CpuTime=21.37 seconds 3 | CPU xRealTime=0.826 or 82.6% of one core 4 | Actual Speech=25.8717 seconds 5 | Utterances=43.84 seconds total 6 | 59% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/result_122.txt: -------------------------------------------------------------------------------- 1 | Utterances=122 2 | CpuTime=177.64 seconds 3 | CPU xRealTime=2.533 or 253.3% of one core 4 | Actual Speech=70.1303 seconds 5 | Utterances=386.37 seconds total 6 | 15% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_pi3/results_66.log: -------------------------------------------------------------------------------- 1 | Utterances=66 2 | CpuTime=140.54 seconds 3 | CPU xRealTime=0.886 or 88.6% of one core 4 | Actual Speech=158.623 seconds 5 | Utterances=251.01 seconds total 6 | 63% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_14.txt: -------------------------------------------------------------------------------- 1 | Utterances=14 2 | CpuTime=62.21 seconds 3 | CPU xRealTime=2.943 or 294.3% of one core 4 | Actual Speech=21.1383 seconds 5 | Utterances=85.15 seconds total 6 | 25% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_18.txt: -------------------------------------------------------------------------------- 1 | Utterances=18 2 | CpuTime=83.26 seconds 3 | CPU xRealTime=2.797 or 279.7% of one core 4 | Actual Speech=29.7676 seconds 5 | Utterances=122.62 seconds total 6 | 24% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_20.txt: -------------------------------------------------------------------------------- 1 | Utterances=20 2 | CpuTime=90.23 seconds 3 | CPU xRealTime=2.872 or 287.2% of one core 4 | Actual Speech=31.4171 seconds 5 | Utterances=127.81 seconds total 6 | 25% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_62.txt: -------------------------------------------------------------------------------- 1 | Utterances=62 2 | CpuTime=206.7 seconds 3 | CPU xRealTime=2.744 or 274.4% of one core 4 | Actual Speech=75.328 seconds 5 | Utterances=326.1 seconds total 6 | 23% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoFile/results/result_PiBplus_file.log: -------------------------------------------------------------------------------- 1 | Utterances=15 2 | CpuTime=219.06 seconds 3 | CPU xRealTime=5.234 or 523.4% of one core 4 | Actual Speech=41.8533 seconds 5 | Utterances=230.94 seconds total 6 | 18% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/results_47.txt: -------------------------------------------------------------------------------- 1 | Utterances=47 2 | CpuTime=202.32 seconds 3 | CPU xRealTime=2.789 or 278.9% of one core 4 | Actual Speech=72.5421 seconds 5 | Utterances=292.98 seconds total 6 | 25% of utterances were speech 7 | -------------------------------------------------------------------------------- /recoFile/input_file.txt: -------------------------------------------------------------------------------- 1 | hello 2 | what time is it 3 | drive forward slowly 4 | who do you think will win the election 5 | what is the weather forecast 6 | how long have you been running 7 | turn forty five degrees left 8 | 9 | one two three four five six seven eight nine ten 10 | 11 | a. b. c. d. e. f. g. h. i j. k. l. m. n. o. p. q. r. s. t. u. v. w. x. y. z. 12 | 13 | goodbye 14 | -------------------------------------------------------------------------------- /recoFile/results/reco_Pi3_file.log: -------------------------------------------------------------------------------- 1 | hello 2 | what time is it 3 | right forward slowly 4 | could you think will win the election 5 | what is the weather forecast 6 | how long have you been running 7 | turn forty five degrees left 8 | 9 | one two three four five six seven eight nine ten 10 | it 11 | a. b. c. d. e. f. g. h. i j. k. l. m. n. o. p. q. r. s. t. e. u. v. w. x. y. z. 12 | 13 | goodbye 14 | 15 | -------------------------------------------------------------------------------- /recoFile/results/reco_PiBplus_file.log: -------------------------------------------------------------------------------- 1 | hello 2 | what time is it 3 | right forward slowly 4 | could you think will win the election 5 | what is the weather forecast 6 | how long have you been running 7 | turn forty five degrees left 8 | 9 | one two three four five six seven eight nine ten 10 | it 11 | a. b. c. d. e. f. g. h. i j. k. l. m. n. o. p. q. r. s. t. e. u. v. w. x. y. z. 12 | 13 | goodbye 14 | 15 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Pi3RoadTest 2 | Speech Recognition using PocketSphinx on Raspberry Pi3 3 | 4 | element14.com awarded me a Raspberry Pi3 for a "Raspberry Pi3 Road Test" 5 | 6 | Short video of each processor running is at: https://youtu.be/dI7jk2sPsLc?si 7 | 8 | This report is located at: https://goo.gl/RrGgCm 9 | 10 | 11 | This repository contains the software and result logs. 12 | 13 | recoMic/ contains recognition test using pocketsphinx with the microphone 14 | 15 | recoFile/ contains recognition test and results for pocketsphinx with file input 16 | 17 | -------------------------------------------------------------------------------- /recoMic/Pi3_Notes.txt: -------------------------------------------------------------------------------- 1 | Notes on Pi3 reco: 2 | 3 | Small LM: 4 | Last four runs: average 82.5% of one core used min: 78% max: 89% 5 | 0 insertion, 0 deletion, 0 substitution = 0% WER 6 | 10 phrase 0.826 xRT 7 | 8 | Lowest seen: 52% 9 | Highest seen: 253% 10 | 11 | Did not change the result if used TTS to play results or not. 12 | Did not change the result if sent output to file. 13 | 14 | 15 | Large LM 10 phrase file: 16 | KiB Mem: 883052 total, 132988 idle 170964 running 17 | 100% CPU 7% memory 18 | 2.168 xRT CPU 19 | 20 | 2 substitutions, 1 deletion, 2 insertions =5 errors / 67 words 21 | = 7% word error rate WER 22 | 23 | 24 | 10 phrase grammar test: 25 | 0 errors -------------------------------------------------------------------------------- /recoMic/PiBplus_Notes.txt: -------------------------------------------------------------------------------- 1 | Notes on Pi B+ reco: 2 | 3 | 2.943 4 | 2.797 5 | 2.872 6 | 2.789 7 | 2.744 8 | --- 9 | 2.829xRT 2.9x max 2.74x min 10 | 11 | 126 phrases in corpus 12 | 13 | 10 phrase small LM test: 9 deletions 4 substitutions 0 additions = 13 errors / 35 words =37%WER 0.45A 90% CPU 5% Memory 3.075 xRT 14 | 15 | 10 phrase large LM test: 16 | top 92-98% CPU 17% memory, 0.40A at 5.02V (+0.08A) 17 | 2 word replacements,1 deletion, 2 insertions = 5 errors / 67 words = 7% Word Error Rate WER 18 | 5.234 xRT reported Total CPU xRT ( sum fwdtree, fwdflat, and bestpath calculations) 19 | 20 | 10 phrase Grammar test: 21 | 1 substitution errors / 35 words = 3% WER 22 | -------------------------------------------------------------------------------- /recoMic/perf.awk: -------------------------------------------------------------------------------- 1 | BEGIN { 2 | cputime = 0.0; 3 | xrealtime = 0.0; 4 | walltime = 0.0; 5 | xwall = 0.0; 6 | utterances = 0; 7 | } 8 | ( $3 == "TOTAL" && $6 == "CPU") { 9 | cputime+=$5; 10 | xrealtime +=$7 } 11 | ( $3 == "TOTAL" && $6 == "wall") { 12 | walltime+=$5; 13 | xwall+=$7; } 14 | ( $3 == "Utterance" ) { 15 | utterances += 1; 16 | } 17 | END { 18 | print "Utterances=" utterances; 19 | print "CpuTime=" cputime " seconds"; 20 | print "CPU xRealTime=" xrealtime " or " xrealtime*100 "% of one core" 21 | print "Actual Speech=" cputime/xrealtime " seconds"; 22 | print "Utterances=" walltime " seconds total"; 23 | printf "%.0f%s of utterances were speech \n", 100/xwall, "%"; 24 | 25 | } 26 | 27 | -------------------------------------------------------------------------------- /recoFile/perf.awk: -------------------------------------------------------------------------------- 1 | BEGIN { 2 | cputime = 0.0; 3 | xrealtime = 0.0; 4 | walltime = 0.0; 5 | xwall = 0.0; 6 | utterances = 0; 7 | } 8 | ( $3 == "TOTAL" && $6 == "CPU") { 9 | cputime+=$5; 10 | xrealtime +=$7 } 11 | ( $3 == "TOTAL" && $6 == "wall") { 12 | walltime+=$5; 13 | xwall+=$7; } 14 | ( $3 == "Utterance" ) { 15 | utterances += 1; 16 | } 17 | END { 18 | print "Utterances=" utterances; 19 | print "CpuTime=" cputime " seconds"; 20 | print "CPU xRealTime=" xrealtime " or " xrealtime*100 "% of one core" 21 | print "Actual Speech=" cputime/xrealtime " seconds"; 22 | print "Utterances=" walltime " seconds total"; 23 | printf "%.0f%s of utterances were speech \n", 100/xwall, "%"; 24 | 25 | } 26 | 27 | -------------------------------------------------------------------------------- /recoMic/pocket_sphinx_listen.py: -------------------------------------------------------------------------------- 1 | # This import will give us our wrapper for the Pocketsphinx library which we can use to get the voice commands from the 2 | # user. 3 | from pocket_sphinx_listener import PocketSphinxListener 4 | import sys 5 | 6 | def runMain(): 7 | # Now we set up the voice recognition using Pocketsphinx from CMU Sphinx. 8 | # We can set debug for the listener here to see messages directly from Pocketsphinx 9 | pocketSphinxListener = PocketSphinxListener(debug=False) 10 | 11 | while True: 12 | try: 13 | # We can set debug here to see what the decoder thinks we are saying as we say it 14 | command = pocketSphinxListener.getCommand(debug=True).lower() 15 | 16 | # Exit when control-c is pressed 17 | except (KeyboardInterrupt, SystemExit): 18 | print 'People sometimes make mistakes, Goodbye.' 19 | sys.exit() 20 | 21 | if __name__ == '__main__': 22 | runMain() 23 | -------------------------------------------------------------------------------- /recoFile/recoFileNotes.txt: -------------------------------------------------------------------------------- 1 | Test Procedure: 2 | 3 | 1) Record test audio of 10 phrases of various length (input_file.txt): 4 | arecord -f s16_LE -r 16000 test16k.wav 5 | Speak: 6 | Hello 7 | What Time is it 8 | Drive Forward Slowly 9 | Who do you think will win the election 10 | What is the weather forecast 11 | How long have you been running 12 | Turn forty five degrees left 13 | one two three four five six seven eight nine ten 14 | a b c d e f g h i j k l m n o p q r s t u v w x y z 15 | Goodbye 16 | 17 | 2) Run large LM PocketSphinx on the recording in one remote ssh, run top in another (%CPU %Mem) 18 | pocketsphinx_continuous -infile test16k.wav 2>&1 | tee ./psphinx.log 19 | (note power consumption A during and after reco) 20 | 21 | 3) Extract Performance Data (xRT) 22 | ./perf.sh >result_Pi_file.log 23 | 24 | 4) Extract recognized phrases 25 | tail -14 >reco_Pi_file.log 26 | 27 | Detailed Test Results: ***** 28 | 29 | 1) pocketsphinx_continuous -infile test16k.wav 30 | Pi B+: top 92-98% CPU 17% memory, 0.40A at 5.02V (+0.08A) 31 | 2 word substitutions, 1 deletion, 2 insertions = 5 errors / 67 words 32 | = 7% Word Error Rate WER 33 | 5.234 xRT reported Total CPU xRT ( sum fwdtree, fwdflat, and bestpath calculations) 34 | Pi 3: top 100% CPU 7% memory, 0.49A (+0.18A) 35 | 2 word substitutions, 1 deletion, 2 insertions = 5 errors / 67 words 36 | = 7% Word Error Rate WER 37 | 2.168 xRT reported Total CPU xRT (sum fwdtree, fwdflat, and bestpath) -------------------------------------------------------------------------------- /recoMic/recoMicNotes.txt: -------------------------------------------------------------------------------- 1 | Test Procedure: 2 | 3 | 4 | 5) Run PocketSphinx from microphone with small LM, speak 10 in-model phrases 5 | python mymain.py 6 | (note power consumption A during and after program) 7 | (note %CPU %Mem from top during program execution) 8 | Speak: 9 | Hello 10 | What Time is it 11 | Drive Forward Slowly 12 | How long have you been running 13 | Turn forty five degrees left 14 | Go backward quickly 15 | Is it going to rain 16 | Spin 17 | Stop now 18 | sudo shutdown minus H now 19 | 20 | 6) Copy term output to LMsmall_Pi.log 21 | 22 | 7) Extract Performance Data (xRT) 23 | ./perf.sh >result_Pi_10.txt 24 | 25 | 8) Run PocketSphinx from microphone with medium JSGF grammar, speak 10 in-grammar phrases. 26 | (note power consumption A during and after program) 27 | (note %CPU %Mem from top during program execution) 28 | Speak: 29 | Hello 30 | What Time is it 31 | Drive Forward Slowly 32 | How long have you been running 33 | Turn forty five degrees left 34 | Go backward quickly 35 | Is it going to rain 36 | Spin 37 | Stop 38 | sudo shutdown minus H now 39 | 40 | 9) Copy term output to jsgf_Pi.log 41 | 42 | 43 | Detailed Test Results: 44 | 45 | 2) pocketsphinx python using microphone and small LM 46 | Pi B+: top 90% CPU 5% memory, 0.45A (+0.13A) 47 | 4 word substitutions, 9 deletion, 0 insertions = 13 errors / 35 words 48 | = 37% Word Error Rate WER 49 | 3.075 xRT reported Total CPU xRT ( sum fwdtree, fwdflat, and bestpath calculations) 50 | Pi 3: top 100% CPU 3% memory, 0.49A (+0.18A) 51 | 0 word substitutions, 0 deletion, 0 insertions = 0 errors / 67 words 52 | = 0% Word Error Rate WER 53 | 0.826 xRT reported Total CPU xRT (sum fwdtree, fwdflat, and bestpath) 54 | 55 | 3) pocketsphinx python using microphone input and medium grammar: 56 | Pi B+: 1 word substitution, 0 deletion, 0 insertions = 1 error / 34 words 57 | = 3% Word Error Rate WER 58 | 59 | Pi 3: 0 word substitutions, 0 deletion, 0 insertions = 0 errors / 34 words 60 | = 0% Word Error Rate WER 61 | -------------------------------------------------------------------------------- /recoMic/mymain.py: -------------------------------------------------------------------------------- 1 | # The following import will allow us to view exceptions with a good level of detail in the case of something unexpected. 2 | import sys, traceback 3 | 4 | # time package has a sleep(seconds) func 5 | import time 6 | 7 | # import subprocess package to run festival tts 8 | import subprocess 9 | 10 | # This import will give us our wrapper for the Pocketsphinx library which we can use to get the voice commands from the 11 | # user. 12 | from pocket_sphinx_listener import PocketSphinxListener 13 | 14 | # Commands in the grammar 15 | # turn on the kitchen light 16 | # turn off the kitchen light 17 | # turn on the bedroom light 18 | # turn off the bedroom light 19 | # turn on the roomba 20 | # turn off the roomba 21 | # roomba clean 22 | # roomba go home 23 | 24 | 25 | def runMyMain(): 26 | 27 | # Now we set up the voice recognition using Pocketsphinx from CMU Sphinx. 28 | pocketSphinxListener = PocketSphinxListener() 29 | 30 | # We want to run forever, or until the user presses control-c, whichever comes first. 31 | while True: 32 | try: 33 | command = pocketSphinxListener.getCommand().lower() 34 | 35 | # for a grammar that looks like TURN 36 | # if command.startswith('turn'): 37 | # onOrOff = command.split()[1] 38 | # deviceName = ''.join(command.split()[2:]) 39 | # do something 40 | # for a grammar that looks like ROOMBA 41 | # elif command.startswith('roomba'): 42 | # action = ' '.join(command.split()[1:]) 43 | # if action == 'clean': 44 | # roomba.clean() 45 | # if action == 'go home': 46 | # roomba.goHome() 47 | # speak what was heard 48 | filename = '_tmp.txt' 49 | file=open(filename,'w') 50 | file.write(command) 51 | file.close() 52 | subprocess.call('festival --tts '+filename, shell=True) 53 | subprocess.call('rm -f '+filename, shell=True) 54 | 55 | # This will allow us to be good cooperators and sleep for a second. 56 | print "I'm thinking now" 57 | time.sleep(1) 58 | 59 | except (KeyboardInterrupt, SystemExit): 60 | print 'Goodbye.' 61 | sys.exit() 62 | except Exception as e: 63 | exc_type, exc_value, exc_traceback = sys.exc_info() 64 | traceback.print_exception(exc_type, exc_value, exc_traceback, 65 | limit=2, 66 | file=sys.stdout) 67 | sys.exit() 68 | 69 | 70 | runMyMain() 71 | 72 | -------------------------------------------------------------------------------- /recoMic/justreco.py: -------------------------------------------------------------------------------- 1 | # The following import will allow us to view exceptions with a good level of detail in the case of something unexpected. 2 | import sys, traceback 3 | 4 | # time package has a sleep(seconds) func 5 | import time 6 | 7 | # import subprocess package to run festival tts 8 | import subprocess 9 | 10 | # This import will give us our wrapper for the Pocketsphinx library which we can use to get the voice commands from the 11 | # user. 12 | from pocket_sphinx_listener import PocketSphinxListener 13 | 14 | # Commands in the grammar 15 | # turn on the kitchen light 16 | # turn off the kitchen light 17 | # turn on the bedroom light 18 | # turn off the bedroom light 19 | # turn on the roomba 20 | # turn off the roomba 21 | # roomba clean 22 | # roomba go home 23 | 24 | 25 | def runMyMain(): 26 | 27 | # Now we set up the voice recognition using Pocketsphinx from CMU Sphinx. 28 | pocketSphinxListener = PocketSphinxListener() 29 | 30 | # We want to run forever, or until the user presses control-c, whichever comes first. 31 | while True: 32 | try: 33 | command = pocketSphinxListener.getCommand().lower() 34 | 35 | # for a grammar that looks like TURN 36 | # if command.startswith('turn'): 37 | # onOrOff = command.split()[1] 38 | # deviceName = ''.join(command.split()[2:]) 39 | # do something 40 | # for a grammar that looks like ROOMBA 41 | # elif command.startswith('roomba'): 42 | # action = ' '.join(command.split()[1:]) 43 | # if action == 'clean': 44 | # roomba.clean() 45 | # if action == 'go home': 46 | # roomba.goHome() 47 | # speak what was heard 48 | # filename = '_tmp.txt' 49 | # file=open(filename,'w') 50 | # file.write(command) 51 | # file.close() 52 | # subprocess.call('festival --tts '+filename, shell=True) 53 | # subprocess.call('rm -f '+filename, shell=True) 54 | 55 | # This will allow us to be good cooperators and sleep for a second. 56 | print "I'm thinking now" 57 | time.sleep(1) 58 | 59 | except (KeyboardInterrupt, SystemExit): 60 | print 'Goodbye.' 61 | sys.exit() 62 | except Exception as e: 63 | exc_type, exc_value, exc_traceback = sys.exc_info() 64 | traceback.print_exception(exc_type, exc_value, exc_traceback, 65 | limit=2, 66 | file=sys.stdout) 67 | sys.exit() 68 | 69 | 70 | runMyMain() 71 | 72 | -------------------------------------------------------------------------------- /recoMic/corpus.txt: -------------------------------------------------------------------------------- 1 | 2 | about face 3 | Alan 4 | Are you listening 5 | Are you on Facebook 6 | at ease 7 | Be quiet 8 | Be very very quiet 9 | Bye 10 | Bye Pi 11 | Bye Pogo 12 | Did you hear me 13 | Did you understand that 14 | Do you do email 15 | Do you have a bed time 16 | Do you have a web site 17 | Do you take dictation 18 | drive 19 | drive forward 20 | drive forward fast 21 | drive forward slowly 22 | Facebook 23 | forward 24 | Go 25 | Go away 26 | Go backward 27 | Go backward slowly 28 | Go fast 29 | Go faster 30 | Go forward 31 | Go forward quickly 32 | Go forward slowly 33 | Goodbye 34 | Good Bye 35 | Go slow 36 | Go slower 37 | Hanna 38 | Hello 39 | Hello Pi 40 | Hello Pogo 41 | Hi 42 | Hi Pi 43 | Hi Pogo 44 | How are you programmed 45 | How long do your batteries last 46 | How long have you been running 47 | How long have you been up 48 | How much disk do you have 49 | How much free disk space 50 | How much free memory 51 | How much memory do you have 52 | How old are you 53 | Is Alan at the computer 54 | Is Alan by the computer 55 | Is Hanna at the computer 56 | Is Hanna in the kitchen 57 | Is it going to rain 58 | Is the sun out 59 | left 60 | Loud 61 | Louder 62 | move back a little 63 | move forward a little 64 | Phone 65 | Pi Alpha Droid 66 | Pogo 67 | Pogo Stop 68 | Quiet 69 | Raspberry Pi 70 | Raspberry Pi B plus 71 | Raspberry Pi Three 72 | Raspberry Pi Two 73 | Record a memo 74 | right 75 | Soft 76 | Softer 77 | Softly 78 | spin 79 | Start 80 | Stop 81 | Stop immediately 82 | Stop now 83 | Sudo reboot 84 | Sudo Shutdown minus H now 85 | Sudo Shutdown minus H plus ten 86 | Sudo Shutdown minus R plus ten 87 | turn forty five degrees left 88 | turn forty five degrees right 89 | Turn left 90 | Turn left a little 91 | Turn left slowly 92 | turn ninety degrees left 93 | turn ninety degrees right 94 | Turn right 95 | Turn right a little 96 | Turn right slowly 97 | What can you do 98 | What day is it 99 | What day is it today 100 | What did you understand 101 | What do you know 102 | What good are you 103 | What is Alan doing 104 | What is Hanna doing 105 | What is the date 106 | What is the date today 107 | What is the time 108 | What is the weather in Boynton Beach 109 | What is your operating system 110 | What is your processor 111 | What kind of robot are you 112 | What languages do you know 113 | What operating system are you running 114 | Whats on TV tonight 115 | What speech recognition package are you running 116 | Whats the weather look like 117 | What text to speech package do you use 118 | What time is it 119 | When do you sleep 120 | When was your last backup 121 | Where do you sleep 122 | Where is Alan 123 | Where is Hanna 124 | Which processor do you have 125 | Whisper 126 | Who made you 127 | -------------------------------------------------------------------------------- /recoMic/results_pi3/mymain.log: -------------------------------------------------------------------------------- 1 | Need more input: 2 | I just heard you say:"YOU LISTENING" 3 | I'm thinking now 4 | Need more input: 5 | I just heard you say:"ARE YOU ON FACEBOOK" 6 | I'm thinking now 7 | Need more input: 8 | I just heard you say:"BE QUIET" 9 | I'm thinking now 10 | Need more input: 11 | I just heard you say:"BE VERY VERY QUIET" 12 | I'm thinking now 13 | Need more input: 14 | I just heard you say:"DID YOU HEAR ME" 15 | I'm thinking now 16 | Need more input: 17 | I just heard you say:"DID YOU UNDERSTAND THAT" 18 | I'm thinking now 19 | Need more input: 20 | I just heard you say:"TAKE DICTATION" 21 | I'm thinking now 22 | Need more input: 23 | I just heard you say:"DRIVE FORWARD SLOWLY" 24 | I'm thinking now 25 | Need more input: 26 | I just heard you say:"GO BACKWARD QUICKLY" 27 | I'm thinking now 28 | Need more input: 29 | I just heard you say:"HELLO POGO" 30 | I'm thinking now 31 | Need more input: 32 | I just heard you say:"HOW ARE YOU PROGRAMMED" 33 | I'm thinking now 34 | Need more input: 35 | I just heard you say:"HOW ARE YOU DO BATTERIES LAST" 36 | I'm thinking now 37 | Need more input: 38 | I just heard you say:"HOW LONG HAVE YOU BEEN RUNNING" 39 | I'm thinking now 40 | Need more input: 41 | I just heard you say:"HOW MUCH FREE MEMORY DO YOU HAVE" 42 | I'm thinking now 43 | Need more input: 44 | I just heard you say:"ALAN AT THE COMPUTER" 45 | I'm thinking now 46 | Need more input: 47 | I just heard you say:"IS HANNA IN THE KITCHEN" 48 | I'm thinking now 49 | Need more input: 50 | I just heard you say:"IS IT GOING TO RAIN" 51 | I'm thinking now 52 | Need more input: 53 | I just heard you say:"IS THE SUN OUT WHAT" 54 | I'm thinking now 55 | Need more input: 56 | I just heard you say:"MOVE BACK A LITTLE WHAT" 57 | I'm thinking now 58 | Need more input: 59 | I just heard you say:"RASPBERRY PI B PLUS" 60 | I'm thinking now 61 | Need more input: 62 | I just heard you say:"RASPBERRY PI FREE" 63 | I'm thinking now 64 | Need more input: 65 | I just heard you say:"IMMEDIATELY" 66 | I'm thinking now 67 | Need more input: 68 | I just heard you say:"GO SHUTDOWN MINUS R NOW" 69 | I'm thinking now 70 | Need more input: 71 | I just heard you say:"SUDO SHUTDOWN MINUS R NOW" 72 | I'm thinking now 73 | Need more input: 74 | I just heard you say:"TURN FORTY FIVE DEGREES LEFT" 75 | I'm thinking now 76 | Need more input: 77 | I just heard you say:"NINETY DEGREES RIGHT" 78 | I'm thinking now 79 | Need more input: 80 | I just heard you say:"WHAT CAN YOU DO" 81 | I'm thinking now 82 | Need more input: 83 | I just heard you say:"WHAT DID YOU UNDERSTAND" 84 | I'm thinking now 85 | Need more input: 86 | I just heard you say:"WHAT IS ALAN DOING" 87 | I'm thinking now 88 | Need more input: 89 | I just heard you say:"WHAT IS THE DATE TODAY" 90 | I'm thinking now 91 | Need more input: 92 | I just heard you say:"WHAT IS YOUR OPERATING SYSTEM" 93 | I'm thinking now 94 | Need more input: 95 | I just heard you say:"WHAT TEXT TO SPEECH PACKAGE DO YOU USE" 96 | I'm thinking now 97 | Need more input: 98 | I just heard you say:"WHEN WHICH YOUR LAST BACKUP" 99 | I'm thinking now 100 | Need more input: 101 | I just heard you say:"ARE" 102 | I'm thinking now 103 | Need more input: 104 | I just heard you say:"ARE" 105 | I'm thinking now 106 | Need more input: 107 | I just heard you say:"WHISPER" 108 | I'm thinking now 109 | Need more input: 110 | I just heard you say:"YOU" 111 | I'm thinking now 112 | Need more input: 113 | Goodbye. 114 | -------------------------------------------------------------------------------- /recoMic/dictionary.dic: -------------------------------------------------------------------------------- 1 | BEDROOM B EH D R UW M 2 | CLEAN K L IY N 3 | GO G OW 4 | HOME HH OW M 5 | KITCHEN K IH CH AH N 6 | LIGHT L AY T 7 | OFF AO F 8 | ON AA N 9 | ON(2) AO N 10 | ROOMBA R UW M B AH 11 | THE DH AH 12 | THE(2) DH IY 13 | TURN T ER N 14 | A AH 15 | A(2) EY 16 | ABOUT AH B AW T 17 | ALAN AE L AH N 18 | ALPHA AE L F AH 19 | ARE AA R 20 | ARE(2) ER 21 | AT AE T 22 | AWAY AH W EY 23 | B B IY 24 | BACK B AE K 25 | BACKUP B AE K AH P 26 | BACKWARD B AE K W ER D 27 | BATTERIES B AE T ER IY Z 28 | BE B IY 29 | BEACH B IY CH 30 | BED B EH D 31 | BEEN B IH N 32 | BEEN(2) B AH N 33 | BOYNTON B OY N T AH N 34 | BYE B AY 35 | CAN K AE N 36 | CAN(2) K AH N 37 | COMPUTER K AH M P Y UW T ER 38 | DATE D EY T 39 | DAY D EY 40 | DEGREES D IH G R IY Z 41 | DICTATION D IH K T EY SH AH N 42 | DID D IH D 43 | DISK D IH S K 44 | DO D UW 45 | DOING D UW IH NG 46 | DRIVE D R AY V 47 | DROID D R OY D 48 | EASE IY Z 49 | EMAIL IY M EY L 50 | FACE F EY S 51 | FACEBOOK F EY S B UH K 52 | FAST F AE S T 53 | FASTER F AE S T ER 54 | FIVE F AY V 55 | FORTY F AO R T IY 56 | FORWARD F AO R W ER D 57 | FREE F R IY 58 | GO G OW 59 | GOING G OW IH NG 60 | GOING(2) G OW IH N 61 | GOOD G UH D 62 | GOOD(2) G IH D 63 | GOODBYE G UH D B AY 64 | H EY CH 65 | HANNA HH AE N AH 66 | HAVE HH AE V 67 | HEAR HH IY R 68 | HELLO HH AH L OW 69 | HELLO(2) HH EH L OW 70 | HOW HH AW 71 | IMMEDIATELY IH M IY D IY AH T L IY 72 | IN IH N 73 | IS IH Z 74 | IT IH T 75 | KIND K AY N D 76 | KITCHEN K IH CH AH N 77 | KNOW N OW 78 | LANGUAGES L AE NG G W AH JH AH Z 79 | LANGUAGES(2) L AE NG G W IH JH IH Z 80 | LAST L AE S T 81 | LEFT L EH F T 82 | LISTENING L IH S AH N IH NG 83 | LISTENING(2) L IH S N IH NG 84 | LITTLE L IH T AH L 85 | LONG L AO NG 86 | LOUDER L AW D ER 87 | MADE M EY D 88 | ME M IY 89 | MEMO M EH M OW 90 | MEMORY M EH M ER IY 91 | MINUS M AY N AH S 92 | MOVE M UW V 93 | MUCH M AH CH 94 | NINETY N AY N T IY 95 | NOW N AW 96 | OF AH V 97 | OLD OW L D 98 | ON AA N 99 | ON(2) AO N 100 | OPERATING AA P ER EY T IH NG 101 | OPERATING(2) AO P ER EY T IH NG 102 | OUT AW T 103 | PACKAGE P AE K AH JH 104 | PACKAGE(2) P AE K IH JH 105 | PHONE F OW N 106 | PI P AY 107 | PLUS P L AH S 108 | POGO P OW G OW 109 | PROCESSOR P R AA S EH S ER 110 | PROGRAMMED P R OW G R AE M D 111 | QUICKLY K W IH K L IY 112 | QUIET K W AY AH T 113 | R AA R 114 | RAIN R EY N 115 | RASPBERRY R AE Z B EH R IY 116 | REBOOT R IY B UW T 117 | RECOGNITION R EH K AH G N IH SH AH N 118 | RECOGNITION(2) R EH K IH G N IH SH AH N 119 | RECORD R AH K AO R D 120 | RECORD(2) R EH K ER D 121 | RECORD(3) R IH K AO R D 122 | RIGHT R AY T 123 | ROBOT R OW B AA T 124 | ROBOT(2) R OW B AH T 125 | RUNNING R AH N IH NG 126 | SHUTDOWN SH AH T D AW N 127 | SITE S AY T 128 | SLEEP S L IY P 129 | SLOW S L OW 130 | SLOWER S L OW ER 131 | SLOWLY S L OW L IY 132 | SPACE S P EY S 133 | SPEECH S P IY CH 134 | SPIN S P IH N 135 | START S T AA R T 136 | STOP S T AA P 137 | SUDO S UW D OW 138 | SUN S AH N 139 | SYSTEM S IH S T AH M 140 | TAKE T EY K 141 | TEN T EH N 142 | TEXT T EH K S T 143 | THAT DH AE T 144 | THAT(2) DH AH T 145 | THE DH AH 146 | THE(2) DH IY 147 | TIME T AY M 148 | TO T UW 149 | TO(2) T IH 150 | TO(3) T AH 151 | TODAY T AH D EY 152 | TODAY(2) T UW D EY 153 | TONIGHT T AH N AY T 154 | TONIGHT(2) T UW N AY T 155 | TURN T ER N 156 | TV T IY V IY 157 | TV(2) T EH L AH V IH ZH AH N 158 | TWO T UW 159 | UNDERSTAND AH N D ER S T AE N D 160 | UP AH P 161 | USE Y UW S 162 | USE(2) Y UW Z 163 | VERY V EH R IY 164 | WAS W AA Z 165 | WEATHER W EH DH ER 166 | WEB W EH B 167 | WHAT W AH T 168 | WHAT(2) HH W AH T 169 | WHATS W AH T S 170 | WHATS(2) HH W AH T S 171 | WHEN W EH N 172 | WHEN(2) HH W EH N 173 | WHEN(3) W IH N 174 | WHEN(4) HH W IH N 175 | WHERE W EH R 176 | WHERE(2) HH W EH R 177 | WHICH W IH CH 178 | WHICH(2) HH W IH CH 179 | WHISPER W IH S P ER 180 | WHISPER(2) HH W IH S P ER 181 | WHO HH UW 182 | YOU Y UW 183 | YOUR Y AO R 184 | YOUR(2) Y UH R 185 | -------------------------------------------------------------------------------- /recoMic/grammar.jsgf: -------------------------------------------------------------------------------- 1 | #JSGF V1.0; 2 | grammar commands; 3 | 4 | 5 | = 6 | ALAN | 7 | HANNA ; 8 | 9 | = 10 | BOYNTON BEACH ; 11 | 12 | = 13 | | 14 | POGO | 15 | [ RASPBERRY ] PI [ ( B PLUS | TWO ) ] | 16 | PI DROID ALPHA | 17 | FACEBOOK ; 18 | 19 | = 20 | LISTENING | 21 | ON FACEBOOK | 22 | UP | 23 | RUNNING ; 24 | 25 | = 26 | HEAR ME | 27 | UNDERSTAND [ ( ME | THAT ) ] ; 28 | 29 | = 30 | A BED TIME | 31 | A WEB SITE ; 32 | 33 | 34 | = 35 | DO EMAIL | 36 | TAKE DICTATION ; 37 | 38 | 39 | = 40 | HOW LONG DO YOUR BATTERIES LAST | 41 | HOW LONG HAVE YOU BEEN ; 42 | 43 | 44 | = 45 | ARE YOU | 46 | DID YOU | 47 | DO YOU HAVE | 48 | DO YOU | 49 | HOW ARE YOU PROGRAMMED | 50 | WHAT DID YOU UNDERSTAND | 51 | WHAT DO YOU KNOW | 52 | WHAT GOOD ARE YOU | 53 | WHAT IS DOING | 54 | WHAT IS THE WEATHER IN | 55 | WHAT IS YOUR OPERATING SYSTEM | 56 | WHAT IS YOUR PROCESSOR | 57 | WHAT KIND OF ROBOT ARE YOU | 58 | WHAT LANGUAGES DO YOU KNOW | 59 | WHAT OPERATING SYSTEM ARE YOU RUNNING | 60 | WHAT SPEECH RECOGNITION PACKAGE ARE YOU RUNNING | 61 | WHAT TEXT TO SPEECH PACKAGE DO YOU USE | 62 | WHATS ON TV TONIGHT | 63 | WHEN DO YOU SLEEP | 64 | WHEN WAS YOUR LAST BACKUP | 65 | WHERE DO YOU SLEEP | 66 | WHERE IS | 67 | WHO MADE YOU ; 68 | 69 | 70 | 71 | = 72 | HOW MUCH [ FREE ] DISK [ SPACE ] DO YOU HAVE | 73 | HOW MUCH [ FREE ] MEMORY DO YOU HAVE | 74 | HOW OLD ARE YOU ; 75 | 76 | 77 | = 78 | [ ( VERY | VERY VERY ) ] QUIET ; 79 | 80 | 81 | = 82 | AWAY | 83 | FORWARD | 84 | BACKWARD | 85 | BACK | 86 | ; 87 | 88 | = 89 | LEFT | 90 | RIGHT ; 91 | 92 | = 93 | FORTY FIVE [ DEGREES ] | 94 | NINETY [ DEGREES ] ; 95 | 96 | 97 | = 98 | BE | 99 | GO | 100 | GO | 101 | STOP | 102 | RECORD A MEMO | 103 | START | 104 | TURN | 105 | DRIVE [ ] | 106 | MOVE [ ] | 107 | SPIN [ ] | 108 | TURN [ ] ; 109 | 110 | = 111 | WHISPER | 112 | ABOUT FACE | 113 | AT EASE ; 114 | 115 | 116 | = 117 | FAST | 118 | SLOW | 119 | SLOWLY | 120 | SLOWER | 121 | FASTER | 122 | QUICKLY | 123 | A LITTLE ; 124 | 125 | = 126 | IMMEDIATELY | 127 | NOW ; 128 | 129 | = 130 | ; 131 | 132 | 133 | = 134 | SUDO SHUTDOWN MINUS H NOW | 135 | SUDO SHUTDOWN MINUS H PLUS TEN | 136 | SUDO SHUTDOWN MINUS R PLUS TEN | 137 | SUDO REBOOT ; 138 | 139 | 140 | = 141 | POGO | 142 | PI ; 143 | 144 | 145 | = 146 | HELLO ; 147 | 148 | = 149 | GOODBYE | 150 | BYE | 151 | GOOD BYE; 152 | 153 | = 154 | ; 155 | 156 | = 157 | ; 158 | 159 | = 160 | COMPUTER | 161 | KITCHEN | 162 | PHONE ; 163 | 164 | = 165 | IS ( AT THE | IN THE ) ; 166 | 167 | 168 | 169 | = 170 | IS IT GOING TO RAIN | 171 | IS THE SUN OUT ; 172 | 173 | = 174 | WHAT CAN YOU DO ; 175 | 176 | = 177 | WHAT ( DAY | DATE ) IS IT [ TODAY ] | 178 | WHAT TIME IS IT [ NOW ] ; 179 | 180 | 181 | 182 | 183 | public = 184 | | 185 | | 186 | | 187 | | 188 | | 189 | | 190 | | 191 | | 192 | | 193 | | 194 | | 195 | | 196 | | 197 | | 198 | ;; 199 | 200 | 201 | 202 | -------------------------------------------------------------------------------- /recoMic/pocket_sphinx_listener.py: -------------------------------------------------------------------------------- 1 | from pocketsphinx import * 2 | import pyaudio 3 | import gevent 4 | 5 | 6 | class PocketSphinxListener(object): 7 | def __init__(self, debug=False): 8 | self.hmm = 'cmusphinx-5prealpha-en-us-ptm-2.0/' 9 | self.dic = 'dictionary.dic' 10 | self.lm = 'language_model.lm' 11 | self.grammar = 'grammar.jsgf' 12 | 13 | self.bitesize = 512 14 | 15 | self.debug = debug 16 | 17 | self.config = Decoder.default_config() 18 | self.config.set_string('-hmm', self.hmm) 19 | # The language model is a statistical model that you can use to determine what words the user is trying to say. 20 | # This can be used in place of a predetermined grammar file. 21 | self.config.set_string('-lm', self.lm) 22 | self.config.set_string('-dict', self.dic) 23 | # self.config.set_string('-jsgf', self.grammar) 24 | # Comment out the following line to get debugging output from the decoder. This is useful if the program is failing 25 | # with an error such as "argument 1 of type 'Decoder *'" 26 | if not self.debug: 27 | self.config.set_string('-logfn', '/dev/null') 28 | # Alan force log 29 | self.config.set_string('-verbose', 'yes') 30 | self.config.set_string('-logfn', 'psphinx.log') 31 | 32 | self.config.set_boolean("-allphone_ci", True) 33 | 34 | self.decoder = Decoder(self.config) 35 | 36 | self.pyAudio = pyaudio.PyAudio() 37 | 38 | def getCommand(self, debug=False): 39 | # Check if we are in debugging mode, either for the getCommand method or for the entire class 40 | if self.debug or debug: 41 | debug = True 42 | 43 | # We're going to set up the stream from pyAudio that well be using to get the user's speech from the microphone. 44 | self.stream = self.pyAudio.open(format=pyaudio.paInt16, 45 | channels=1, 46 | rate=16000, 47 | input=True, 48 | frames_per_buffer=self.bitesize) 49 | 50 | # Let the use know that we're ready for them to speak 51 | print 'Need more input: ' 52 | 53 | # This is a flag that we'll use in a bit to determine whether we are going from silence to speech 54 | # or from speech to silence. 55 | utteranceStarted = False 56 | 57 | # This will tell PocketSphinx to start decoding the "utterance". When we are finished with our audio 58 | # we will tell PocketSphinx that the utterance is over. 59 | self.decoder.start_utt() 60 | 61 | # We want this to loop for as long as it takes to get the full sentence from the user. We only exit with a 62 | # return statement when we have our best guess of what the person said. 63 | while True: 64 | try: 65 | # This takes a small sound bite from the microphone to process. 66 | soundBite = self.stream.read(self.bitesize) 67 | except Exception as e: 68 | pass 69 | 70 | # If we've got something from the microphone, we should begin processing it. 71 | if soundBite: 72 | self.decoder.process_raw(soundBite, False, False) 73 | inSpeech = self.decoder.get_in_speech() 74 | # The following checks for the transition from silence to speech. 75 | # We're going to set a flag to reflect this. 76 | if inSpeech and not utteranceStarted: 77 | utteranceStarted = True 78 | # The following checks for the transition from speech to silence. 79 | # This is our cue to check what was said and do something useful with it. 80 | if not inSpeech and utteranceStarted: 81 | # We tell PocketSphinx that the user is finished saying what they wanted 82 | # to say, and that it should makes it's best guess as to what thay was. 83 | self.decoder.end_utt() 84 | # The following will get a hypothesis object with, amongst other things, 85 | # the string of words that PocketSphinx thinks the user said. 86 | self.hypothesis = self.decoder.hyp() 87 | if self.hypothesis is not None: 88 | bestGuess = self.hypothesis.hypstr 89 | print 'I just heard you say:"{}"'.format(bestGuess) 90 | # We are done with the microphone for now so we'll close the stream. 91 | self.stream.stop_stream() 92 | self.stream.close() 93 | # We have what we came for! A string representing what the user said. 94 | # We'll now return it to the runMain function so that it can be 95 | # processed and some meaning can be gleamed from it. 96 | return bestGuess 97 | # The following is here for debugging to see what the decoder thinks we're saying as we go 98 | if debug and self.decoder.hyp() is not None: 99 | print self.decoder.hyp().hypstr 100 | -------------------------------------------------------------------------------- /recoMic/pocket_sphinx_listener.gram.py: -------------------------------------------------------------------------------- 1 | from pocketsphinx import * 2 | import pyaudio 3 | import gevent 4 | 5 | 6 | class PocketSphinxListener(object): 7 | def __init__(self, debug=False): 8 | self.hmm = 'cmusphinx-5prealpha-en-us-ptm-2.0/' 9 | self.dic = 'dictionary.dic' 10 | self.lm = 'language_model.lm' 11 | self.grammar = 'grammar.jsgf' 12 | 13 | self.bitesize = 512 14 | 15 | self.debug = debug 16 | 17 | self.config = Decoder.default_config() 18 | self.config.set_string('-hmm', self.hmm) 19 | # The language model is a statistical model that you can use to determine what words the user is trying to say. 20 | # This can be used in place of a predetermined grammar file. 21 | # self.config.set_string('-lm', self.lm) 22 | self.config.set_string('-dict', self.dic) 23 | self.config.set_string('-jsgf', self.grammar) 24 | # Comment out the following line to get debugging output from the decoder. This is useful if the program is failing 25 | # with an error such as "argument 1 of type 'Decoder *'" 26 | if not self.debug: 27 | self.config.set_string('-logfn', '/dev/null') 28 | # Alan force log 29 | self.config.set_string('-verbose', 'yes') 30 | self.config.set_string('-logfn', 'psphinx.log') 31 | 32 | self.config.set_boolean("-allphone_ci", True) 33 | 34 | self.decoder = Decoder(self.config) 35 | 36 | self.pyAudio = pyaudio.PyAudio() 37 | 38 | def getCommand(self, debug=False): 39 | # Check if we are in debugging mode, either for the getCommand method or for the entire class 40 | if self.debug or debug: 41 | debug = True 42 | 43 | # We're going to set up the stream from pyAudio that well be using to get the user's speech from the microphone. 44 | self.stream = self.pyAudio.open(format=pyaudio.paInt16, 45 | channels=1, 46 | rate=16000, 47 | input=True, 48 | frames_per_buffer=self.bitesize) 49 | 50 | # Let the use know that we're ready for them to speak 51 | print 'Need more input: ' 52 | 53 | # This is a flag that we'll use in a bit to determine whether we are going from silence to speech 54 | # or from speech to silence. 55 | utteranceStarted = False 56 | 57 | # This will tell PocketSphinx to start decoding the "utterance". When we are finished with our audio 58 | # we will tell PocketSphinx that the utterance is over. 59 | self.decoder.start_utt() 60 | 61 | # We want this to loop for as long as it takes to get the full sentence from the user. We only exit with a 62 | # return statement when we have our best guess of what the person said. 63 | while True: 64 | try: 65 | # This takes a small sound bite from the microphone to process. 66 | soundBite = self.stream.read(self.bitesize) 67 | except Exception as e: 68 | pass 69 | 70 | # If we've got something from the microphone, we should begin processing it. 71 | if soundBite: 72 | self.decoder.process_raw(soundBite, False, False) 73 | inSpeech = self.decoder.get_in_speech() 74 | # The following checks for the transition from silence to speech. 75 | # We're going to set a flag to reflect this. 76 | if inSpeech and not utteranceStarted: 77 | utteranceStarted = True 78 | # The following checks for the transition from speech to silence. 79 | # This is our cue to check what was said and do something useful with it. 80 | if not inSpeech and utteranceStarted: 81 | # We tell PocketSphinx that the user is finished saying what they wanted 82 | # to say, and that it should makes it's best guess as to what thay was. 83 | self.decoder.end_utt() 84 | # The following will get a hypothesis object with, amongst other things, 85 | # the string of words that PocketSphinx thinks the user said. 86 | self.hypothesis = self.decoder.hyp() 87 | if self.hypothesis is not None: 88 | bestGuess = self.hypothesis.hypstr 89 | print 'I just heard you say:"{}"'.format(bestGuess) 90 | # We are done with the microphone for now so we'll close the stream. 91 | self.stream.stop_stream() 92 | self.stream.close() 93 | # We have what we came for! A string representing what the user said. 94 | # We'll now return it to the runMain function so that it can be 95 | # processed and some meaning can be gleamed from it. 96 | return bestGuess 97 | # The following is here for debugging to see what the decoder thinks we're saying as we go 98 | if debug and self.decoder.hyp() is not None: 99 | print self.decoder.hyp().hypstr 100 | -------------------------------------------------------------------------------- /recoMic/pocket_sphinx_listener.lm.py: -------------------------------------------------------------------------------- 1 | from pocketsphinx import * 2 | import pyaudio 3 | import gevent 4 | 5 | 6 | class PocketSphinxListener(object): 7 | def __init__(self, debug=False): 8 | self.hmm = 'cmusphinx-5prealpha-en-us-ptm-2.0/' 9 | self.dic = 'dictionary.dic' 10 | self.lm = 'language_model.lm' 11 | self.grammar = 'grammar.jsgf' 12 | 13 | self.bitesize = 512 14 | 15 | self.debug = debug 16 | 17 | self.config = Decoder.default_config() 18 | self.config.set_string('-hmm', self.hmm) 19 | # The language model is a statistical model that you can use to determine what words the user is trying to say. 20 | # This can be used in place of a predetermined grammar file. 21 | self.config.set_string('-lm', self.lm) 22 | self.config.set_string('-dict', self.dic) 23 | # self.config.set_string('-jsgf', self.grammar) 24 | # Comment out the following line to get debugging output from the decoder. This is useful if the program is failing 25 | # with an error such as "argument 1 of type 'Decoder *'" 26 | if not self.debug: 27 | self.config.set_string('-logfn', '/dev/null') 28 | # Alan force log 29 | self.config.set_string('-verbose', 'yes') 30 | self.config.set_string('-logfn', 'psphinx.log') 31 | 32 | self.config.set_boolean("-allphone_ci", True) 33 | 34 | self.decoder = Decoder(self.config) 35 | 36 | self.pyAudio = pyaudio.PyAudio() 37 | 38 | def getCommand(self, debug=False): 39 | # Check if we are in debugging mode, either for the getCommand method or for the entire class 40 | if self.debug or debug: 41 | debug = True 42 | 43 | # We're going to set up the stream from pyAudio that well be using to get the user's speech from the microphone. 44 | self.stream = self.pyAudio.open(format=pyaudio.paInt16, 45 | channels=1, 46 | rate=16000, 47 | input=True, 48 | frames_per_buffer=self.bitesize) 49 | 50 | # Let the use know that we're ready for them to speak 51 | print 'Need more input: ' 52 | 53 | # This is a flag that we'll use in a bit to determine whether we are going from silence to speech 54 | # or from speech to silence. 55 | utteranceStarted = False 56 | 57 | # This will tell PocketSphinx to start decoding the "utterance". When we are finished with our audio 58 | # we will tell PocketSphinx that the utterance is over. 59 | self.decoder.start_utt() 60 | 61 | # We want this to loop for as long as it takes to get the full sentence from the user. We only exit with a 62 | # return statement when we have our best guess of what the person said. 63 | while True: 64 | try: 65 | # This takes a small sound bite from the microphone to process. 66 | soundBite = self.stream.read(self.bitesize) 67 | except Exception as e: 68 | pass 69 | 70 | # If we've got something from the microphone, we should begin processing it. 71 | if soundBite: 72 | self.decoder.process_raw(soundBite, False, False) 73 | inSpeech = self.decoder.get_in_speech() 74 | # The following checks for the transition from silence to speech. 75 | # We're going to set a flag to reflect this. 76 | if inSpeech and not utteranceStarted: 77 | utteranceStarted = True 78 | # The following checks for the transition from speech to silence. 79 | # This is our cue to check what was said and do something useful with it. 80 | if not inSpeech and utteranceStarted: 81 | # We tell PocketSphinx that the user is finished saying what they wanted 82 | # to say, and that it should makes it's best guess as to what thay was. 83 | self.decoder.end_utt() 84 | # The following will get a hypothesis object with, amongst other things, 85 | # the string of words that PocketSphinx thinks the user said. 86 | self.hypothesis = self.decoder.hyp() 87 | if self.hypothesis is not None: 88 | bestGuess = self.hypothesis.hypstr 89 | print 'I just heard you say:"{}"'.format(bestGuess) 90 | # We are done with the microphone for now so we'll close the stream. 91 | self.stream.stop_stream() 92 | self.stream.close() 93 | # We have what we came for! A string representing what the user said. 94 | # We'll now return it to the runMain function so that it can be 95 | # processed and some meaning can be gleamed from it. 96 | return bestGuess 97 | # The following is here for debugging to see what the decoder thinks we're saying as we go 98 | if debug and self.decoder.hyp() is not None: 99 | print self.decoder.hyp().hypstr 100 | -------------------------------------------------------------------------------- /Sphinx On Pi3.txt: -------------------------------------------------------------------------------- 1 | 2 | 3 | makezine.com Roomba, I Command Thee: Use Raspberry Pi for Voice Control 4 | http://makezine.com/projects/use-raspberry-pi-for-voice-control/ 5 | 6 | 7 | 8 | First, go get the packages required for SphinxBase by executing: 9 | 10 | sudo apt-get update 11 | 12 | sudo apt-get install libasound2-dev autoconf libtool bison \ 13 | 14 | swig python-dev python-pyaudio 15 | 16 | You’ll also need to install some Python libraries for use with our demo application. To do this, you’ll install and use the Python pip command with the following commands: 17 | 18 | curl -O https://bootstrap.pypa.io/get-pip.py 19 | 20 | sudo python get-pip.py 21 | 22 | sudo pip install gevent grequests 23 | 24 | TIP: If your connection to the Pi is a bit flaky and prone to disconnects, you can save yourself some heartache by running these commands in a screen session. To do so, run the following before continuing. 25 | 26 | sudo apt-get install screen 27 | 28 | screen -DR sphinx 29 | 30 | If at any stage you get disconnected from your Pi (and it hasn’t restarted) you can run screen -DR sphinx again to reconnect and continue where you left off. 31 | 32 | OBTAINING THE SPHINX TOOLS 33 | Now you can go about getting the SphinxBase package, which is used by PocketSphinx as well as other software in the CMU Sphinx family. 34 | 35 | To obtain SphinxBase execute the following commands: 36 | 37 | git clone git://github.com/cmusphinx/sphinxbase.git 38 | 39 | cd sphinxbase 40 | 41 | git checkout 3b34d87 42 | 43 | ./autogen.sh 44 | 45 | make 46 | 47 | (At this stage you may want to go make coffee …) 48 | 49 | sudo make install 50 | 51 | cd .. 52 | 53 | You’re ready to move on to PocketSphinx. To obtain PocketSphinx, execute the following commands: 54 | 55 | git clone git://github.com/cmusphinx/pocketsphinx.git 56 | 57 | cd pocketsphinx 58 | 59 | git checkout 4e4e607 60 | 61 | ./autogen.sh 62 | 63 | make 64 | 65 | (Time for a second cup of coffee …) 66 | 67 | sudo make install 68 | 69 | cd .. 70 | 71 | To update the system with your new libraries, run sudo ldconfig. 72 | 73 | TESTING THE SPEECH RECOGNITION 74 | Now that you have the building blocks of your speech recognition in place, you’ll want to test that it actually works before continuing. 75 | 76 | Now you can run a test of PocketSphinx using pocketsphinx_continuous -inmic yes. 77 | 78 | You should see something like the following, which indicates the system is ready for you to start speaking: 79 | 80 | Listening... 81 | 82 | Input overrun, read calls are too rare (non-fatal) 83 | 84 | You can safely ignore the warning. Go ahead and speak! 85 | 86 | When you’re finished, you should see some technical information along with PocketSphinx’s best guess as to what you said, and then another READY prompt letting you know it’s ready for more input. 87 | 88 | INFO: ngram_search.c(874): bestpath 0.10 CPU 0.071 xRT 89 | 90 | INFO: ngram_search.c(877): bestpath 0.11 wall 0.078 xRT 91 | 92 | what 93 | 94 | READY.... 95 | 96 | RECO FROM FILE with large LM: 97 | arecord -f s16_LE -r 16000 test16k.wav 98 | pocketsphinx_continuous -infile test16k.wav 2>&1 | tee ./psphinx.log 99 | 100 | xRT= sum of fwdflat, CPU xRT 101 | 102 | --CONTROL ALL THE THINGS-- 103 | For our demo application, I’ve programmed our system to be able to control three separate systems: Philips Hue and Insteon lighting systems, and an iRobot Roomba robot vacuum cleaner. ... 104 | 105 | ..retrieve the Python source code: 106 | 107 | git clone https://github.com/bynds/makevoicedemo 108 | 109 | ---- 110 | 111 | USING POCKETSPHINX 112 | There are several modes that you can configure for PocketSphinx. For example, it can be asked to listen for a specific keyword (it will attempt to ignore everything it hears except the keyword), or it can be asked to use a grammar that you specify (it will try to fit everything it hears into the confines of the grammar). We are using the grammar mode in our example, with a grammar that’s designed to allow us to capture all the commands we’ll be using. The grammar file is specified in JSGF or JSpeech Grammar Format which has a powerful yet straightforward syntax for specifying the speech that it expects to hear in terms of simple rules. 113 | 114 | In addition to the grammar file, you’re going to need three more files in order to use PocketSphinx in our application: a dictionary file which will define words in terms of how they sound, a language model file which contains statistics about the words and their order, and an acoustic model which is used to determine how audio correlates with the sounds in words. The grammar file, dictionary, and language model will all be generated specifically for our project, while the acoustic model will be a generic model for U.S. English. 115 | 116 | GENERATING THE DICTIONARY 117 | In order to generate our dictionary, we will be making use of lmtool, the web based tool hosted by CMU specifically for quickly generating these files. The input to lmtool is a corpus file which contains all or most of the sentences that you would like to be able to recognize. In our simple use case, we have the following sentences in our corpus: 118 | 119 | turn on the kitchen light 120 | turn off the kitchen light 121 | turn on the bedroom light 122 | turn off the bedroom light 123 | turn on the roomba 124 | turn off the roomba 125 | roomba clean 126 | roomba go home 127 | 128 | You can type these into a text editor and save the file as corpus.txt or you can download a readymade version from the Github repository. 129 | 130 | Now that you have your corpus file, go use lmtool. To upload your corpus file, click the Browse button which will bring up a dialog box that allows you to select the corpus file you just created. 131 | 132 | Then click the button Compile Knowledge Base. You’ll be taken to a page with links to download the result. You can either download the compressed .tgz file which contains all the files generated or simply download the .dic file labeled Pronunciation Dictionary. Copy this file to the same makevoicedemo directory that was created on the Pi earlier. You can rename the file using the command mv *.dic dictionary.dic to make it easier to work with. 133 | 134 | While you’re at it, download the prebuilt acoustic model from the Sphinx Sourceforge. Once you’ve moved it to the makevoicedemo directory, extract it with: 135 | 136 | tar -xvf cmusphinx-en-us-ptm-5.2.tar.gz. 137 | 138 | CREATING THE GRAMMAR FILE 139 | As I mentioned earlier, everything that PocketSphinx hears, it will try and fit into the words of the grammar. Check out how the JSGF format is described in the W3C note. It starts with a declaration of the format followed by a declaration of the grammar name. We simply called ours “commands.” 140 | 141 | We have chosen to use three main rules: an action, an object, and a command. For each rule, you’ll define “tokens” which are what you expect the user to say. For example, the two tokens for our action rule are TURN ON and TURN OFF. We therefore represent the rule as: 142 | 143 | 144 | 145 | = TURN ON | 146 | 147 | TURN OFF ; 148 | 149 | 150 | Similarly the _object_ rule we define as: 151 | 152 | = KITCHEN LIGHT| 153 | 154 | BEDROOM LIGHT| 155 | 156 | ROOMBA ; 157 | 158 | 159 | Finally, to demonstrate that we can nest rules or create them with explicit tokens, we define a command as: 160 | 161 | public = THE | 162 | 163 | ROOMBA CLEAN | 164 | 165 | ROOMBA GO HOME ; 166 | 167 | 168 | Notice the public keyword in front of the . This allows us to use the rule by importing it into other grammar files in the future. 169 | 170 | INITIALIZING THE DECODER 171 | We are using Python as our programming language because it is easy to read, powerful, and thanks to the foresight of the PocketSphinx developers, it’s also very easy to use with PocketSphinx. 172 | 173 | The main workhorse when recognizing speech with PocketSphinx is the decoder. In order to use the decoder we must first set a config for the decoder to use. 174 | 175 | from pocketsphinx import * 176 | 177 | hmm = 'cmusphinx-5prealpha-en-us-ptm-2.0/' 178 | 179 | dic = 'dictionary.dic' 180 | 181 | grammar = 'grammar.jsgf' 182 | 183 | config = Decoder.default_config() 184 | 185 | config.set_string('-hmm', hmm) 186 | 187 | config.set_string('-dict', dic) 188 | 189 | config.set_string('-jsgf', grammar) 190 | 191 | Once this is done, initializing a decoder is as simple as decoder = Decoder(config). 192 | 193 | For the example application, we’re using the pyAudio library to get the user’s speech from the microphone for processing by PocketSphinx. The specifics of this library are less important for our purposes (investigating speech recognition) and we will therefore simply take it for granted that pyAudio works as advertised. 194 | 195 | The specifics of obtaining the decoder’s text output are a bit complex, however the basic process can be distilled down to the following steps. 196 | 197 | 198 | \# Start an 'utterance' 199 | 200 | decoder.start_utt() 201 | 202 | \# Process a soundbite 203 | 204 | decoder.process_raw(soundBite, False, False) 205 | 206 | \# End the utterance when the user finishes speaking 207 | 208 | decoder.end_utt() 209 | 210 | \# Retrieve the hypothesis (for what was said) 211 | 212 | hypothesis = decoder.hyp() 213 | 214 | \# Get the text of the hypothesis 215 | 216 | bestGuess = hypothesis.hypstr 217 | 218 | \# Print out what was said 219 | 220 | print 'I just heard you say:"{}"'.format(bestGuess) 221 | 222 | Those interested in learning more about the gritty details of this process should turn their attention to the pocketSphinxListener.py code from the example project. 223 | 224 | There are a lot of different configuration parameters that you can experiment with, and as previously mentioned, other modes of recognition to try. For instance, investigate the -allphone_ci PocketSphinx configuration option and its impact on decoding accuracy. Or try keyword spotting for activating a light. Or try a statistical language model, like the one that was generated when we were using the lmtool earlier, instead of a grammar file. As a practitioner you can experiment almost endlessly to explore the fringes of what’s possible. One thing you’ll quickly notice is that PocketSphinx is an actively developed research system and this will sometimes mean you need to rewrite your application to match the new APIs and function names. 225 | 226 | 227 | ====== alan ===== 228 | sudo apt-get install alsa-tools alsa-utils 229 | 230 | sudo nano /usr/share/alsa/alsa.conf 231 | (replace all lines like pcm.front cards.pcm.front <-- pcm.front cards.pcm.default ) 232 | 233 | ... 234 | pcm.front cards.pcm.default 235 | pcm.rear cards.pcm.default 236 | pcm.center_lfe cards.pcm.default 237 | pcm.side cards.pcm.default 238 | pcm.surround21 cards.pcm.default 239 | pcm.surround40 cards.pcm.default 240 | pcm.surround41 cards.pcm.default 241 | pcm.surround50 cards.pcm.default 242 | pcm.surround51 cards.pcm.default 243 | pcm.surround71 cards.pcm.default 244 | pcm.iec958 cards.pcm.default 245 | pcm.spdif cards.pcm.default 246 | pcm.hdmi cards.pcm.default 247 | pcm.dmix cards.pcm.default 248 | pcm.dsnoop cards.pcm.default 249 | pcm.modem cards.pcm.default 250 | pcm.phoneline cards.pcm.default 251 | 252 | ... 253 | (cntl-o,enter,cntl-x) 254 | 255 | then 256 | sudo alsactl init 257 | 258 | ========= mymain.py ======== 259 | # The following import will allow us to view exceptions with a good level of detail in the case of something unexpected. 260 | import sys, traceback 261 | 262 | # time package has a sleep(seconds) func 263 | import time 264 | 265 | # import subprocess package to run festival tts 266 | import subprocess 267 | 268 | # This import will give us our wrapper for the Pocketsphinx library which we can use to get the voice commands from the 269 | # user. 270 | from pocket_sphinx_listener import PocketSphinxListener 271 | 272 | # Commands in the grammar 273 | # turn on the kitchen light 274 | # turn off the kitchen light 275 | # turn on the bedroom light 276 | # turn off the bedroom light 277 | # turn on the roomba 278 | # turn off the roomba 279 | # roomba clean 280 | # roomba go home 281 | 282 | 283 | def runMyMain(): 284 | 285 | # Now we set up the voice recognition using Pocketsphinx from CMU Sphinx. 286 | pocketSphinxListener = PocketSphinxListener() 287 | 288 | # We want to run forever, or until the user presses control-c, whichever comes first. 289 | while True: 290 | try: 291 | command = pocketSphinxListener.getCommand().lower() 292 | 293 | # for a grammar that looks like TURN 294 | # if command.startswith('turn'): 295 | # onOrOff = command.split()[1] 296 | # deviceName = ''.join(command.split()[2:]) 297 | # do something 298 | # for a grammar that looks like ROOMBA 299 | # elif command.startswith('roomba'): 300 | # action = ' '.join(command.split()[1:]) 301 | # if action == 'clean': 302 | # roomba.clean() 303 | # if action == 'go home': 304 | # roomba.goHome() 305 | 306 | 307 | # speak what was heard 308 | filename = '_tmp.txt' 309 | file=open(filename,'w') 310 | file.write(command) 311 | file.close() 312 | subprocess.call('festival --tts '+filename, shell=True) 313 | subprocess.call('rm -f '+filename, shell=True) 314 | 315 | # This will allow us to be good cooperators and sleep for a second. 316 | print "I'm thinking now" 317 | time.sleep(1) 318 | 319 | except (KeyboardInterrupt, SystemExit): 320 | print 'Goodbye.' 321 | sys.exit() 322 | except Exception as e: 323 | exc_type, exc_value, exc_traceback = sys.exc_info() 324 | traceback.print_exception(exc_type, exc_value, exc_traceback, 325 | limit=2, 326 | file=sys.stdout) 327 | sys.exit() 328 | 329 | 330 | runMyMain() 331 | ================ 332 | 333 | In lm mode log shows performance information (latest psphinx supposedly will show for grammar mode also). 334 | 335 | For individual or TOTAL: 336 | Add up fwdtree + fwdflat + bestpath CPU time = CPU time spent recognizing 337 | Add up fwdtree + fwdflat + bestpath xRT (<1 e.g. 0.52 means 1s of audio takes 0.52 seconds of CPU time, or 0.52% of one core to perform the reco. 338 | (To calculate length of audio processed divide total CPU time by percent of CPU) 339 | 340 | on Pi 3: (126 phrase corpus using 136 words, 278 bi-grams, 295 tri-grams) 341 | 1.2GHz single core processing 342 | 343 | 64 phrases: 344 | Total CPU: 75.7s 0.52 xRT 146s audio 345 | Total Wall: 190s 1.3 xRT (146s audio) - 346 | Wall time includes startup, tear down, output and logging. 347 | Reco from mic cannot be faster than realtime. 348 | 0.52 xRT CPU means 52% of one core used by ASR. 349 | 350 | 351 | 352 | 353 | 354 | 355 | 356 | 357 | -------------------------------------------------------------------------------- /recoMic/language_model.lm: -------------------------------------------------------------------------------- 1 | Language model created by QuickLM on Sat Aug 8 21:15:11 EDT 2015 2 | Copyright (c) 1996-2010 Carnegie Mellon University and Alexander I. Rudnicky 3 | 4 | The model is in standard ARPA format, designed by Doug Paul while he was at MITRE. 5 | 6 | The code that was used to produce this language model is available in Open Source. 7 | Please visit http://www.speech.cs.cmu.edu/tools/ for more information 8 | 9 | The (fixed) discount mass is 0.5. The backoffs are computed using the ratio method. 10 | This model based on a corpus of 106 sentences and 136 words 11 | 12 | \data\ 13 | ngram 1=136 14 | ngram 2=278 15 | ngram 3=295 16 | 17 | \1-grams: 18 | -1.0421 -0.3010 19 | -1.0421 -0.2327 20 | -2.2223 A -0.2984 21 | -3.0674 ABOUT -0.3007 22 | -2.4654 ALAN -0.2577 23 | -3.0674 ALPHA -0.3007 24 | -2.1644 ARE -0.2909 25 | -2.5903 AT -0.2977 26 | -3.0674 AWAY -0.2597 27 | -3.0674 B -0.2999 28 | -3.0674 BACK -0.2984 29 | -3.0674 BACKUP -0.2597 30 | -2.7664 BACKWARD -0.2577 31 | -3.0674 BATTERIES -0.3003 32 | -2.7664 BE -0.2992 33 | -3.0674 BEACH -0.2597 34 | -3.0674 BED -0.2999 35 | -2.7664 BEEN -0.2995 36 | -3.0674 BOYNTON -0.3007 37 | -3.0674 CAN -0.2909 38 | -2.7664 COMPUTER -0.2597 39 | -2.7664 DATE -0.2589 40 | -2.7664 DAY -0.2943 41 | -2.4654 DEGREES -0.2965 42 | -3.0674 DICTATION -0.2597 43 | -2.5903 DID -0.2909 44 | -2.7664 DISK -0.2950 45 | -1.8914 DO -0.2464 46 | -2.7664 DOING -0.2597 47 | -2.5903 DRIVE -0.2568 48 | -3.0674 DROID -0.2597 49 | -3.0674 EASE -0.2597 50 | -3.0674 EMAIL -0.2597 51 | -3.0674 FACE -0.2597 52 | -2.7664 FACEBOOK -0.2597 53 | -3.0674 FAST -0.2597 54 | -2.7664 FIVE -0.2995 55 | -2.7664 FORTY -0.3003 56 | -2.2223 FORWARD -0.2544 57 | -2.7664 FREE -0.2995 58 | -2.1132 GO -0.2548 59 | -3.0674 GOING -0.3003 60 | -3.0674 GOOD -0.2980 61 | -2.7664 H -0.2992 62 | -2.3685 HANNA -0.2568 63 | -2.2223 HAVE -0.2456 64 | -3.0674 HEAR -0.3007 65 | -2.7664 HELLO -0.2585 66 | -2.1132 HOW -0.2950 67 | -3.0674 IMMEDIATELY -0.2597 68 | -2.7664 IN -0.2977 69 | -1.8122 IS -0.2916 70 | -2.4654 IT -0.2585 71 | -3.0674 KIND -0.3007 72 | -3.0674 KITCHEN -0.2597 73 | -2.7664 KNOW -0.2597 74 | -3.0674 LANGUAGES -0.2954 75 | -2.7664 LAST -0.2593 76 | -2.2893 LEFT -0.2548 77 | -3.0674 LISTENING -0.2597 78 | -2.4654 LITTLE -0.2597 79 | -2.5903 LONG -0.2928 80 | -3.0674 LOUDER -0.2597 81 | -3.0674 MADE -0.2909 82 | -3.0674 ME -0.2597 83 | -3.0674 MEMO -0.2597 84 | -2.7664 MEMORY -0.2535 85 | -2.5903 MINUS -0.2999 86 | -2.7664 MOVE -0.2980 87 | -2.4654 MUCH -0.2988 88 | -2.7664 NINETY -0.2995 89 | -2.7664 NOW -0.2597 90 | -3.0674 OF -0.3007 91 | -3.0674 OLD -0.2980 92 | -2.7664 ON -0.2999 93 | -2.7664 OPERATING -0.3003 94 | -3.0674 OUT -0.2597 95 | -2.7664 PACKAGE -0.2924 96 | -3.0674 PHONE -0.2597 97 | -2.4654 PI -0.2585 98 | -2.5903 PLUS -0.2589 99 | -2.5903 POGO -0.2581 100 | -2.7664 PROCESSOR -0.2535 101 | -3.0674 PROGRAMMED -0.2597 102 | -3.0674 QUICKLY -0.2597 103 | -2.5903 QUIET -0.2597 104 | -3.0674 R -0.2999 105 | -3.0674 RAIN -0.2597 106 | -2.5903 RASPBERRY -0.2995 107 | -3.0674 REBOOT -0.2597 108 | -3.0674 RECOGNITION -0.3003 109 | -3.0674 RECORD -0.2984 110 | -2.2893 RIGHT -0.2548 111 | -3.0674 ROBOT -0.2980 112 | -2.5903 RUNNING -0.2597 113 | -2.5903 SHUTDOWN -0.2999 114 | -3.0674 SITE -0.2597 115 | -2.7664 SLEEP -0.2597 116 | -3.0674 SLOW -0.2597 117 | -2.3685 SLOWLY -0.2597 118 | -3.0674 SPACE -0.2597 119 | -2.7664 SPEECH -0.2999 120 | -3.0674 SPIN -0.2597 121 | -3.0674 START -0.2597 122 | -2.4654 STOP -0.2585 123 | -2.4654 SUDO -0.2995 124 | -3.0674 SUN -0.3007 125 | -2.7664 SYSTEM -0.2564 126 | -3.0674 TAKE -0.3007 127 | -2.7664 TEN -0.2597 128 | -3.0674 TEXT -0.3003 129 | -3.0674 THAT -0.2597 130 | -2.1644 THE -0.2973 131 | -2.5903 TIME -0.2523 132 | -2.7664 TO -0.2999 133 | -2.7664 TODAY -0.2597 134 | -3.0674 TONIGHT -0.2597 135 | -2.0674 TURN -0.2950 136 | -3.0674 TV -0.3007 137 | -3.0674 TWO -0.2597 138 | -2.7664 UNDERSTAND -0.2593 139 | -3.0674 UP -0.2597 140 | -3.0674 USE -0.2597 141 | -2.7664 VERY -0.2992 142 | -3.0674 WAS -0.2995 143 | -3.0674 WEATHER -0.3003 144 | -3.0674 WEB -0.3007 145 | -1.7664 WHAT -0.2820 146 | -3.0674 WHATS -0.3003 147 | -2.7664 WHEN -0.2950 148 | -2.5903 WHERE -0.2886 149 | -3.0674 WHICH -0.3003 150 | -3.0674 WHISPER -0.2597 151 | -3.0674 WHO -0.3007 152 | -1.6361 YOU -0.2430 153 | -2.4654 YOUR -0.2984 154 | 155 | \2-grams: 156 | -2.3263 ABOUT 0.0000 157 | -2.3263 ALAN -0.1761 158 | -2.0253 ARE 0.0000 159 | -2.3263 AT -0.2218 160 | -2.0253 BE 0.0000 161 | -2.0253 DID 0.0000 162 | -1.7243 DO -0.0792 163 | -1.8492 DRIVE 0.0000 164 | -2.3263 FACEBOOK 0.0000 165 | -2.3263 FORWARD -0.1963 166 | -1.3721 GO 0.0000 167 | -2.3263 HANNA -0.2041 168 | -2.0253 HELLO 0.0000 169 | -1.3721 HOW 0.0000 170 | -1.6274 IS -0.0458 171 | -2.3263 LEFT -0.1249 172 | -2.3263 LOUDER 0.0000 173 | -2.0253 MOVE 0.0000 174 | -2.3263 PHONE 0.0000 175 | -2.3263 PI -0.2430 176 | -2.0253 POGO 0.0000 177 | -2.3263 QUIET 0.0000 178 | -1.8492 RASPBERRY 0.0000 179 | -2.3263 RECORD 0.0000 180 | -2.3263 RIGHT -0.1249 181 | -2.3263 SPIN 0.0000 182 | -2.3263 START 0.0000 183 | -1.8492 STOP 0.0000 184 | -1.7243 SUDO 0.0000 185 | -1.3263 TURN 0.0000 186 | -1.0253 WHAT 0.0000 187 | -2.3263 WHATS 0.0000 188 | -2.0253 WHEN 0.0000 189 | -1.8492 WHERE 0.0000 190 | -2.3263 WHICH 0.0000 191 | -2.3263 WHISPER 0.0000 192 | -2.3263 WHO 0.0000 193 | -1.1461 A BED 0.0000 194 | -0.5441 A LITTLE 0.0000 195 | -1.1461 A MEMO 0.0000 196 | -1.1461 A WEB 0.0000 197 | -0.3010 ABOUT FACE 0.0000 198 | -0.6021 ALAN -0.3010 199 | -0.9031 ALAN AT -0.1249 200 | -0.9031 ALAN DOING 0.0000 201 | -0.3010 ALPHA DROID 0.0000 202 | -0.3010 ARE YOU -0.2218 203 | -0.7782 AT EASE 0.0000 204 | -0.4771 AT THE -0.2430 205 | -0.3010 AWAY -0.3010 206 | -0.3010 B PLUS -0.2218 207 | -0.3010 BACK A -0.1549 208 | -0.3010 BACKUP -0.3010 209 | -0.6021 BACKWARD -0.3010 210 | -0.6021 BACKWARD SLOWLY 0.0000 211 | -0.3010 BATTERIES LAST -0.1761 212 | -0.6021 BE QUIET 0.0000 213 | -0.6021 BE VERY -0.1761 214 | -0.3010 BEACH -0.3010 215 | -0.3010 BED TIME -0.1249 216 | -0.6021 BEEN RUNNING 0.0000 217 | -0.6021 BEEN UP 0.0000 218 | -0.3010 BOYNTON BEACH 0.0000 219 | -0.3010 CAN YOU -0.2846 220 | -0.3010 COMPUTER -0.3010 221 | -0.6021 DATE -0.3010 222 | -0.6021 DATE TODAY 0.0000 223 | -0.3010 DAY IS -0.2499 224 | -0.6021 DEGREES LEFT -0.1249 225 | -0.6021 DEGREES RIGHT -0.1249 226 | -0.3010 DICTATION -0.3010 227 | -0.3010 DID YOU -0.2762 228 | -0.6021 DISK DO -0.0792 229 | -0.6021 DISK SPACE 0.0000 230 | -1.4771 DO -0.3010 231 | -1.4771 DO EMAIL 0.0000 232 | -0.3979 DO YOU -0.1814 233 | -1.4771 DO YOUR -0.2430 234 | -0.3010 DOING -0.3010 235 | -0.7782 DRIVE -0.3010 236 | -0.4771 DRIVE FORWARD -0.1091 237 | -0.3010 DROID -0.3010 238 | -0.3010 EASE -0.3010 239 | -0.3010 EMAIL -0.3010 240 | -0.3010 FACE -0.3010 241 | -0.3010 FACEBOOK -0.3010 242 | -0.3010 FAST -0.3010 243 | -0.3010 FIVE DEGREES 0.0000 244 | -0.3010 FORTY FIVE 0.0000 245 | -0.6690 FORWARD -0.3010 246 | -1.1461 FORWARD A -0.1549 247 | -1.1461 FORWARD QUICKLY 0.0000 248 | -0.8451 FORWARD SLOWLY 0.0000 249 | -0.6021 FREE DISK -0.1761 250 | -0.6021 FREE MEMORY -0.1761 251 | -1.2553 GO -0.3010 252 | -1.2553 GO AWAY 0.0000 253 | -0.9542 GO BACKWARD 0.0000 254 | -1.2553 GO FAST 0.0000 255 | -0.7782 GO FORWARD -0.0580 256 | -1.2553 GO SLOW 0.0000 257 | -0.3010 GOING TO -0.1761 258 | -0.3010 GOOD ARE 0.0000 259 | -0.6021 H NOW 0.0000 260 | -0.6021 H PLUS -0.1249 261 | -0.6990 HANNA -0.3010 262 | -1.0000 HANNA AT -0.1249 263 | -1.0000 HANNA DOING 0.0000 264 | -1.0000 HANNA IN -0.1761 265 | -0.6690 HAVE -0.3010 266 | -0.8451 HAVE A -0.2341 267 | -0.8451 HAVE YOU -0.2846 268 | -0.3010 HEAR ME 0.0000 269 | -0.6021 HELLO -0.3010 270 | -0.6021 HELLO POGO -0.1249 271 | -1.2553 HOW ARE 0.0000 272 | -0.7782 HOW LONG 0.0000 273 | -0.6532 HOW MUCH 0.0000 274 | -1.2553 HOW OLD 0.0000 275 | -0.3010 IMMEDIATELY -0.3010 276 | -0.6021 IN BOYNTON 0.0000 277 | -0.6021 IN THE -0.2730 278 | -1.0792 IS ALAN 0.0000 279 | -0.9542 IS HANNA 0.0000 280 | -0.9542 IS IT 0.0000 281 | -0.8573 IS THE -0.1383 282 | -1.2553 IS YOUR -0.1761 283 | -0.6021 IT -0.3010 284 | -0.9031 IT GOING 0.0000 285 | -0.9031 IT TODAY 0.0000 286 | -0.3010 KIND OF 0.0000 287 | -0.3010 KITCHEN -0.3010 288 | -0.3010 KNOW -0.3010 289 | -0.3010 LANGUAGES DO -0.0792 290 | -0.6021 LAST -0.3010 291 | -0.6021 LAST BACKUP 0.0000 292 | -0.4771 LEFT -0.3010 293 | -1.0792 LEFT A -0.1549 294 | -1.0792 LEFT SLOWLY 0.0000 295 | -0.3010 LISTENING -0.3010 296 | -0.3010 LITTLE -0.3010 297 | -0.7782 LONG DO -0.2863 298 | -0.4771 LONG HAVE -0.2341 299 | -0.3010 LOUDER -0.3010 300 | -0.3010 MADE YOU -0.2676 301 | -0.3010 ME -0.3010 302 | -0.3010 MEMO -0.3010 303 | -0.6021 MEMORY -0.3010 304 | -0.6021 MEMORY DO -0.0792 305 | -0.4771 MINUS H 0.0000 306 | -0.7782 MINUS R 0.0000 307 | -0.6021 MOVE BACK 0.0000 308 | -0.6021 MOVE FORWARD -0.2688 309 | -0.9031 MUCH DISK -0.1761 310 | -0.6021 MUCH FREE 0.0000 311 | -0.9031 MUCH MEMORY -0.1761 312 | -0.3010 NINETY DEGREES 0.0000 313 | -0.3010 NOW -0.3010 314 | -0.3010 OF ROBOT 0.0000 315 | -0.3010 OLD ARE 0.0000 316 | -0.6021 ON FACEBOOK 0.0000 317 | -0.6021 ON TV 0.0000 318 | -0.3010 OPERATING SYSTEM 0.0000 319 | -0.3010 OUT -0.3010 320 | -0.6021 PACKAGE ARE 0.0000 321 | -0.6021 PACKAGE DO -0.0792 322 | -0.3010 PHONE -0.3010 323 | -0.9031 PI -0.3010 324 | -0.9031 PI ALPHA 0.0000 325 | -0.9031 PI B 0.0000 326 | -0.9031 PI TWO 0.0000 327 | -0.7782 PLUS -0.3010 328 | -0.4771 PLUS TEN 0.0000 329 | -0.4771 POGO -0.3010 330 | -0.7782 POGO STOP -0.1761 331 | -0.6021 PROCESSOR -0.3010 332 | -0.6021 PROCESSOR DO -0.0792 333 | -0.3010 PROGRAMMED -0.3010 334 | -0.3010 QUICKLY -0.3010 335 | -0.3010 QUIET -0.3010 336 | -0.3010 R PLUS -0.1249 337 | -0.3010 RAIN -0.3010 338 | -0.3010 RASPBERRY PI -0.0969 339 | -0.3010 REBOOT -0.3010 340 | -0.3010 RECOGNITION PACKAGE -0.1761 341 | -0.3010 RECORD A -0.2688 342 | -0.4771 RIGHT -0.3010 343 | -1.0792 RIGHT A -0.1549 344 | -1.0792 RIGHT SLOWLY 0.0000 345 | -0.3010 ROBOT ARE 0.0000 346 | -0.3010 RUNNING -0.3010 347 | -0.3010 SHUTDOWN MINUS 0.0000 348 | -0.3010 SITE -0.3010 349 | -0.3010 SLEEP -0.3010 350 | -0.3010 SLOW -0.3010 351 | -0.3010 SLOWLY -0.3010 352 | -0.3010 SPACE -0.3010 353 | -0.6021 SPEECH PACKAGE -0.1761 354 | -0.6021 SPEECH RECOGNITION 0.0000 355 | -0.3010 SPIN -0.3010 356 | -0.3010 START -0.3010 357 | -0.6021 STOP -0.3010 358 | -0.9031 STOP IMMEDIATELY 0.0000 359 | -0.9031 STOP NOW 0.0000 360 | -0.9031 SUDO REBOOT 0.0000 361 | -0.4260 SUDO SHUTDOWN 0.0000 362 | -0.3010 SUN OUT 0.0000 363 | -0.6021 SYSTEM -0.3010 364 | -0.6021 SYSTEM ARE 0.0000 365 | -0.3010 TAKE DICTATION 0.0000 366 | -0.3010 TEN -0.3010 367 | -0.3010 TEXT TO -0.1761 368 | -0.3010 THAT -0.3010 369 | -0.9031 THE COMPUTER 0.0000 370 | -0.9031 THE DATE 0.0000 371 | -1.2041 THE KITCHEN 0.0000 372 | -1.2041 THE SUN 0.0000 373 | -1.2041 THE TIME -0.1249 374 | -1.2041 THE WEATHER 0.0000 375 | -0.4771 TIME -0.3010 376 | -0.7782 TIME IS -0.2499 377 | -0.6021 TO RAIN 0.0000 378 | -0.6021 TO SPEECH -0.1761 379 | -0.3010 TODAY -0.3010 380 | -0.3010 TONIGHT -0.3010 381 | -1.0000 TURN FORTY 0.0000 382 | -0.8239 TURN LEFT 0.0000 383 | -1.0000 TURN NINETY 0.0000 384 | -0.8239 TURN RIGHT 0.0000 385 | -0.3010 TV TONIGHT 0.0000 386 | -0.3010 TWO -0.3010 387 | -0.6021 UNDERSTAND -0.3010 388 | -0.6021 UNDERSTAND THAT 0.0000 389 | -0.3010 UP -0.3010 390 | -0.3010 USE -0.3010 391 | -0.6021 VERY QUIET 0.0000 392 | -0.6021 VERY VERY -0.1761 393 | -0.3010 WAS YOUR -0.2430 394 | -0.3010 WEATHER IN -0.1761 395 | -0.3010 WEB SITE 0.0000 396 | -1.6021 WHAT CAN 0.0000 397 | -1.3010 WHAT DAY 0.0000 398 | -1.6021 WHAT DID 0.0000 399 | -1.6021 WHAT DO -0.0792 400 | -1.6021 WHAT GOOD 0.0000 401 | -0.6990 WHAT IS -0.0872 402 | -1.6021 WHAT KIND 0.0000 403 | -1.6021 WHAT LANGUAGES 0.0000 404 | -1.6021 WHAT OPERATING 0.0000 405 | -1.6021 WHAT SPEECH -0.1761 406 | -1.6021 WHAT TEXT 0.0000 407 | -1.6021 WHAT TIME -0.2218 408 | -0.3010 WHATS ON -0.1761 409 | -0.6021 WHEN DO -0.0792 410 | -0.6021 WHEN WAS 0.0000 411 | -0.7782 WHERE DO -0.0792 412 | -0.4771 WHERE IS -0.2071 413 | -0.3010 WHICH PROCESSOR -0.1761 414 | -0.3010 WHISPER -0.3010 415 | -0.3010 WHO MADE 0.0000 416 | -1.1303 YOU -0.3010 417 | -1.4314 YOU BEEN 0.0000 418 | -1.4314 YOU DO -0.2711 419 | -1.0334 YOU HAVE -0.1091 420 | -1.7324 YOU HEAR 0.0000 421 | -1.4314 YOU KNOW 0.0000 422 | -1.7324 YOU LISTENING 0.0000 423 | -1.7324 YOU ON -0.1761 424 | -1.7324 YOU PROGRAMMED 0.0000 425 | -1.4314 YOU RUNNING 0.0000 426 | -1.4314 YOU SLEEP 0.0000 427 | -1.7324 YOU TAKE 0.0000 428 | -1.4314 YOU UNDERSTAND 0.0000 429 | -1.7324 YOU USE 0.0000 430 | -0.9031 YOUR BATTERIES 0.0000 431 | -0.9031 YOUR LAST -0.1761 432 | -0.9031 YOUR OPERATING 0.0000 433 | -0.9031 YOUR PROCESSOR -0.1761 434 | 435 | \3-grams: 436 | -0.3010 ABOUT FACE 437 | -0.3010 ALAN 438 | -0.3010 ARE YOU 439 | -0.3010 AT EASE 440 | -0.6021 BE QUIET 441 | -0.6021 BE VERY 442 | -0.3010 DID YOU 443 | -0.3010 DO YOU 444 | -0.7782 DRIVE 445 | -0.4771 DRIVE FORWARD 446 | -0.3010 FACEBOOK 447 | -0.3010 FORWARD 448 | -1.2553 GO 449 | -1.2553 GO AWAY 450 | -0.9542 GO BACKWARD 451 | -1.2553 GO FAST 452 | -0.7782 GO FORWARD 453 | -1.2553 GO SLOW 454 | -0.3010 HANNA 455 | -0.6021 HELLO 456 | -0.6021 HELLO POGO 457 | -1.2553 HOW ARE 458 | -0.7782 HOW LONG 459 | -0.6532 HOW MUCH 460 | -1.2553 HOW OLD 461 | -1.0000 IS ALAN 462 | -0.6990 IS HANNA 463 | -1.0000 IS IT 464 | -1.0000 IS THE 465 | -0.3010 LEFT 466 | -0.3010 LOUDER 467 | -0.6021 MOVE BACK 468 | -0.6021 MOVE FORWARD 469 | -0.3010 PHONE 470 | -0.3010 PI ALPHA 471 | -0.6021 POGO 472 | -0.6021 POGO STOP 473 | -0.3010 QUIET 474 | -0.3010 RASPBERRY PI 475 | -0.3010 RECORD A 476 | -0.3010 RIGHT 477 | -0.3010 SPIN 478 | -0.3010 START 479 | -0.7782 STOP 480 | -0.7782 STOP IMMEDIATELY 481 | -0.7782 STOP NOW 482 | -0.9031 SUDO REBOOT 483 | -0.4260 SUDO SHUTDOWN 484 | -1.0000 TURN FORTY 485 | -0.8239 TURN LEFT 486 | -1.0000 TURN NINETY 487 | -0.8239 TURN RIGHT 488 | -1.6021 WHAT CAN 489 | -1.3010 WHAT DAY 490 | -1.6021 WHAT DID 491 | -1.6021 WHAT DO 492 | -1.6021 WHAT GOOD 493 | -0.6990 WHAT IS 494 | -1.6021 WHAT KIND 495 | -1.6021 WHAT LANGUAGES 496 | -1.6021 WHAT OPERATING 497 | -1.6021 WHAT SPEECH 498 | -1.6021 WHAT TEXT 499 | -1.6021 WHAT TIME 500 | -0.3010 WHATS ON 501 | -0.6021 WHEN DO 502 | -0.6021 WHEN WAS 503 | -0.7782 WHERE DO 504 | -0.4771 WHERE IS 505 | -0.3010 WHICH PROCESSOR 506 | -0.3010 WHISPER 507 | -0.3010 WHO MADE 508 | -0.3010 A BED TIME 509 | -0.3010 A LITTLE 510 | -0.3010 A MEMO 511 | -0.3010 A WEB SITE 512 | -0.3010 ABOUT FACE 513 | -0.3010 ALAN AT THE 514 | -0.3010 ALAN DOING 515 | -0.3010 ALPHA DROID 516 | -0.7270 ARE YOU 517 | -1.2041 ARE YOU LISTENING 518 | -1.2041 ARE YOU ON 519 | -1.2041 ARE YOU PROGRAMMED 520 | -0.9031 ARE YOU RUNNING 521 | -0.3010 AT EASE 522 | -0.3010 AT THE COMPUTER 523 | -0.3010 B PLUS 524 | -0.3010 BACK A LITTLE 525 | -0.3010 BACKWARD SLOWLY 526 | -0.3010 BATTERIES LAST 527 | -0.3010 BE QUIET 528 | -0.3010 BE VERY VERY 529 | -0.3010 BED TIME 530 | -0.3010 BEEN RUNNING 531 | -0.3010 BEEN UP 532 | -0.3010 BOYNTON BEACH 533 | -0.3010 CAN YOU DO 534 | -0.3010 DATE TODAY 535 | -0.3010 DAY IS IT 536 | -0.3010 DEGREES LEFT 537 | -0.3010 DEGREES RIGHT 538 | -0.7782 DID YOU HEAR 539 | -0.4771 DID YOU UNDERSTAND 540 | -0.3010 DISK DO YOU 541 | -0.3010 DISK SPACE 542 | -0.3010 DO EMAIL 543 | -1.3802 DO YOU DO 544 | -0.6812 DO YOU HAVE 545 | -1.0792 DO YOU KNOW 546 | -1.0792 DO YOU SLEEP 547 | -1.3802 DO YOU TAKE 548 | -1.3802 DO YOU USE 549 | -0.3010 DO YOUR BATTERIES 550 | -0.6021 DRIVE FORWARD 551 | -0.6021 DRIVE FORWARD SLOWLY 552 | -0.6021 FIVE DEGREES LEFT 553 | -0.6021 FIVE DEGREES RIGHT 554 | -0.3010 FORTY FIVE DEGREES 555 | -0.3010 FORWARD A LITTLE 556 | -0.3010 FORWARD QUICKLY 557 | -0.3010 FORWARD SLOWLY 558 | -0.3010 FREE DISK SPACE 559 | -0.3010 FREE MEMORY 560 | -0.3010 GO AWAY 561 | -0.6021 GO BACKWARD 562 | -0.6021 GO BACKWARD SLOWLY 563 | -0.3010 GO FAST 564 | -0.7782 GO FORWARD 565 | -0.7782 GO FORWARD QUICKLY 566 | -0.7782 GO FORWARD SLOWLY 567 | -0.3010 GO SLOW 568 | -0.3010 GOING TO RAIN 569 | -0.3010 GOOD ARE YOU 570 | -0.3010 H NOW 571 | -0.3010 H PLUS TEN 572 | -0.3010 HANNA AT THE 573 | -0.3010 HANNA DOING 574 | -0.3010 HANNA IN THE 575 | -0.6021 HAVE A BED 576 | -0.6021 HAVE A WEB 577 | -0.3010 HAVE YOU BEEN 578 | -0.3010 HEAR ME 579 | -0.3010 HELLO POGO 580 | -0.3010 HOW ARE YOU 581 | -0.7782 HOW LONG DO 582 | -0.4771 HOW LONG HAVE 583 | -0.9031 HOW MUCH DISK 584 | -0.6021 HOW MUCH FREE 585 | -0.9031 HOW MUCH MEMORY 586 | -0.3010 HOW OLD ARE 587 | -0.3010 IN BOYNTON BEACH 588 | -0.3010 IN THE KITCHEN 589 | -0.7782 IS ALAN 590 | -0.7782 IS ALAN AT 591 | -0.7782 IS ALAN DOING 592 | -0.9031 IS HANNA 593 | -0.9031 IS HANNA AT 594 | -0.9031 IS HANNA DOING 595 | -0.9031 IS HANNA IN 596 | -0.6021 IS IT 597 | -0.9031 IS IT GOING 598 | -0.9031 IS IT TODAY 599 | -0.6990 IS THE DATE 600 | -1.0000 IS THE SUN 601 | -1.0000 IS THE TIME 602 | -1.0000 IS THE WEATHER 603 | -0.6021 IS YOUR OPERATING 604 | -0.6021 IS YOUR PROCESSOR 605 | -0.3010 IT GOING TO 606 | -0.3010 IT TODAY 607 | -0.3010 KIND OF ROBOT 608 | -0.3010 LANGUAGES DO YOU 609 | -0.3010 LAST BACKUP 610 | -0.3010 LEFT A LITTLE 611 | -0.3010 LEFT SLOWLY 612 | -0.3010 LONG DO YOUR 613 | -0.3010 LONG HAVE YOU 614 | -0.3010 MADE YOU 615 | -0.3010 MEMORY DO YOU 616 | -0.6021 MINUS H NOW 617 | -0.6021 MINUS H PLUS 618 | -0.3010 MINUS R PLUS 619 | -0.3010 MOVE BACK A 620 | -0.3010 MOVE FORWARD A 621 | -0.3010 MUCH DISK DO 622 | -0.6021 MUCH FREE DISK 623 | -0.6021 MUCH FREE MEMORY 624 | -0.3010 MUCH MEMORY DO 625 | -0.6021 NINETY DEGREES LEFT 626 | -0.6021 NINETY DEGREES RIGHT 627 | -0.3010 OF ROBOT ARE 628 | -0.3010 OLD ARE YOU 629 | -0.3010 ON FACEBOOK 630 | -0.3010 ON TV TONIGHT 631 | -0.6021 OPERATING SYSTEM 632 | -0.6021 OPERATING SYSTEM ARE 633 | -0.3010 PACKAGE ARE YOU 634 | -0.3010 PACKAGE DO YOU 635 | -0.3010 PI ALPHA DROID 636 | -0.3010 PI B PLUS 637 | -0.3010 PI TWO 638 | -0.3010 PLUS TEN 639 | -0.3010 POGO STOP 640 | -0.3010 PROCESSOR DO YOU 641 | -0.3010 R PLUS TEN 642 | -0.7782 RASPBERRY PI 643 | -0.7782 RASPBERRY PI B 644 | -0.7782 RASPBERRY PI TWO 645 | -0.3010 RECOGNITION PACKAGE ARE 646 | -0.3010 RECORD A MEMO 647 | -0.3010 RIGHT A LITTLE 648 | -0.3010 RIGHT SLOWLY 649 | -0.3010 ROBOT ARE YOU 650 | -0.4771 SHUTDOWN MINUS H 651 | -0.7782 SHUTDOWN MINUS R 652 | -0.3010 SPEECH PACKAGE DO 653 | -0.3010 SPEECH RECOGNITION PACKAGE 654 | -0.3010 STOP IMMEDIATELY 655 | -0.3010 STOP NOW 656 | -0.3010 SUDO REBOOT 657 | -0.3010 SUDO SHUTDOWN MINUS 658 | -0.3010 SUN OUT 659 | -0.3010 SYSTEM ARE YOU 660 | -0.3010 TAKE DICTATION 661 | -0.3010 TEXT TO SPEECH 662 | -0.3010 THE COMPUTER 663 | -0.6021 THE DATE 664 | -0.6021 THE DATE TODAY 665 | -0.3010 THE KITCHEN 666 | -0.3010 THE SUN OUT 667 | -0.3010 THE TIME 668 | -0.3010 THE WEATHER IN 669 | -0.3010 TIME IS IT 670 | -0.3010 TO RAIN 671 | -0.3010 TO SPEECH PACKAGE 672 | -0.3010 TURN FORTY FIVE 673 | -0.7782 TURN LEFT 674 | -0.7782 TURN LEFT A 675 | -0.7782 TURN LEFT SLOWLY 676 | -0.3010 TURN NINETY DEGREES 677 | -0.7782 TURN RIGHT 678 | -0.7782 TURN RIGHT A 679 | -0.7782 TURN RIGHT SLOWLY 680 | -0.3010 TV TONIGHT 681 | -0.3010 UNDERSTAND THAT 682 | -0.3010 VERY QUIET 683 | -0.3010 VERY VERY QUIET 684 | -0.3010 WAS YOUR LAST 685 | -0.3010 WEATHER IN BOYNTON 686 | -0.3010 WEB SITE 687 | -0.3010 WHAT CAN YOU 688 | -0.3010 WHAT DAY IS 689 | -0.3010 WHAT DID YOU 690 | -0.3010 WHAT DO YOU 691 | -0.3010 WHAT GOOD ARE 692 | -1.2041 WHAT IS ALAN 693 | -1.2041 WHAT IS HANNA 694 | -0.6021 WHAT IS THE 695 | -0.9031 WHAT IS YOUR 696 | -0.3010 WHAT KIND OF 697 | -0.3010 WHAT LANGUAGES DO 698 | -0.3010 WHAT OPERATING SYSTEM 699 | -0.3010 WHAT SPEECH RECOGNITION 700 | -0.3010 WHAT TEXT TO 701 | -0.3010 WHAT TIME IS 702 | -0.3010 WHATS ON TV 703 | -0.3010 WHEN DO YOU 704 | -0.3010 WHEN WAS YOUR 705 | -0.3010 WHERE DO YOU 706 | -0.6021 WHERE IS ALAN 707 | -0.6021 WHERE IS HANNA 708 | -0.3010 WHICH PROCESSOR DO 709 | -0.3010 WHO MADE YOU 710 | -0.6021 YOU BEEN RUNNING 711 | -0.6021 YOU BEEN UP 712 | -0.6021 YOU DO 713 | -0.6021 YOU DO EMAIL 714 | -0.5229 YOU HAVE 715 | -0.6990 YOU HAVE A 716 | -0.3010 YOU HEAR ME 717 | -0.3010 YOU KNOW 718 | -0.3010 YOU LISTENING 719 | -0.3010 YOU ON FACEBOOK 720 | -0.3010 YOU PROGRAMMED 721 | -0.3010 YOU RUNNING 722 | -0.3010 YOU SLEEP 723 | -0.3010 YOU TAKE DICTATION 724 | -0.6021 YOU UNDERSTAND 725 | -0.6021 YOU UNDERSTAND THAT 726 | -0.3010 YOU USE 727 | -0.3010 YOUR BATTERIES LAST 728 | -0.3010 YOUR LAST BACKUP 729 | -0.3010 YOUR OPERATING SYSTEM 730 | -0.3010 YOUR PROCESSOR 731 | 732 | \end\ 733 | -------------------------------------------------------------------------------- /recoMic/results_pi3/results.log: -------------------------------------------------------------------------------- 1 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.70 CPU 0.376 xRT 2 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.44 wall 1.314 xRT 3 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.253 xRT 4 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.47 wall 0.253 xRT 5 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 6 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.006 xRT 7 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.01 CPU 0.365 xRT 8 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.40 wall 1.229 xRT 9 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.30 CPU 0.108 xRT 10 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.30 wall 0.107 xRT 11 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 12 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 13 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.63 CPU 0.375 xRT 14 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.35 wall 1.398 xRT 15 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.36 CPU 0.214 xRT 16 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.217 xRT 17 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 18 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 19 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.78 CPU 0.379 xRT 20 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.86 wall 1.389 xRT 21 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.41 CPU 0.199 xRT 22 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.203 xRT 23 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 24 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 25 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.76 CPU 0.362 xRT 26 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.18 wall 1.039 xRT 27 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.162 xRT 28 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.163 xRT 29 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 30 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 31 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.48 CPU 0.375 xRT 32 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.39 wall 1.086 xRT 33 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.19 CPU 0.148 xRT 34 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.19 wall 0.145 xRT 35 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 36 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 37 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.08 CPU 0.361 xRT 38 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.05 wall 1.021 xRT 39 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.45 CPU 0.151 xRT 40 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.45 wall 0.149 xRT 41 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 42 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 43 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.75 CPU 0.366 xRT 44 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.034 xRT 45 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.132 xRT 46 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.27 wall 0.133 xRT 47 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 48 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 49 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.59 CPU 0.351 xRT 50 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.74 wall 1.035 xRT 51 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.155 xRT 52 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.25 wall 0.150 xRT 53 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 54 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 55 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.75 CPU 0.359 xRT 56 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.18 wall 1.045 xRT 57 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.105 xRT 58 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.22 wall 0.107 xRT 59 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 60 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 61 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.27 CPU 0.422 xRT 62 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.71 wall 1.114 xRT 63 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.13 CPU 0.203 xRT 64 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.13 wall 0.210 xRT 65 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 66 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.002 xRT 67 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.73 CPU 0.372 xRT 68 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.09 wall 1.067 xRT 69 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.32 CPU 0.163 xRT 70 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.163 xRT 71 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 72 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 73 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.66 CPU 0.361 xRT 74 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.90 wall 1.037 xRT 75 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.35 CPU 0.191 xRT 76 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.195 xRT 77 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 78 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 79 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.69 CPU 0.381 xRT 80 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.82 wall 1.559 xRT 81 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.31 CPU 0.171 xRT 82 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.31 wall 0.170 xRT 83 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 84 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 85 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.35 CPU 0.422 xRT 86 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.83 wall 2.204 xRT 87 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.10 CPU 0.120 xRT 88 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.10 wall 0.116 xRT 89 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 90 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 91 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.36 CPU 0.419 xRT 92 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.93 wall 1.085 xRT 93 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.15 CPU 0.174 xRT 94 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.15 wall 0.176 xRT 95 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.012 xRT 96 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 97 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.61 CPU 0.367 xRT 98 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.73 wall 1.044 xRT 99 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.18 CPU 0.108 xRT 100 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.18 wall 0.109 xRT 101 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 102 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 103 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.67 CPU 0.356 xRT 104 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.96 wall 1.041 xRT 105 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.117 xRT 106 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.22 wall 0.117 xRT 107 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 108 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 109 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.74 CPU 0.354 xRT 110 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.16 wall 1.032 xRT 111 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.19 CPU 0.091 xRT 112 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.19 wall 0.093 xRT 113 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 114 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 115 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.14 CPU 0.344 xRT 116 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.40 wall 1.027 xRT 117 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.142 xRT 118 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.140 xRT 119 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 120 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 121 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.54 CPU 0.380 xRT 122 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.495 xRT 123 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.24 CPU 0.169 xRT 124 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.23 wall 0.164 xRT 125 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 126 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 127 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.92 CPU 0.376 xRT 128 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.27 wall 1.334 xRT 129 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.31 CPU 0.127 xRT 130 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.31 wall 0.128 xRT 131 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 132 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 133 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.76 CPU 0.376 xRT 134 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.73 wall 1.351 xRT 135 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.134 xRT 136 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.28 wall 0.138 xRT 137 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 138 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 139 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.35 CPU 0.366 xRT 140 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.86 wall 1.589 xRT 141 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.37 CPU 0.100 xRT 142 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.37 wall 0.099 xRT 143 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 144 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 145 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.56 CPU 0.368 xRT 146 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.58 wall 1.039 xRT 147 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.21 CPU 0.138 xRT 148 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.20 wall 0.134 xRT 149 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 150 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 151 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.80 CPU 0.374 xRT 152 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.079 xRT 153 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.38 CPU 0.178 xRT 154 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.38 wall 0.176 xRT 155 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 156 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.003 xRT 157 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.62 CPU 0.371 xRT 158 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.16 wall 1.291 xRT 159 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.156 xRT 160 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.26 wall 0.157 xRT 161 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 162 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 163 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.84 CPU 0.368 xRT 164 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.35 wall 1.029 xRT 165 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.50 CPU 0.219 xRT 166 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.50 wall 0.218 xRT 167 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 168 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.003 xRT 169 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.98 CPU 0.344 xRT 170 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.99 wall 1.048 xRT 171 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.42 CPU 0.147 xRT 172 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.146 xRT 173 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 174 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 175 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.48 CPU 0.360 xRT 176 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.17 wall 1.015 xRT 177 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.33 CPU 0.080 xRT 178 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.083 xRT 179 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.002 xRT 180 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 181 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.13 CPU 0.370 xRT 182 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.23 wall 1.387 xRT 183 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.38 CPU 0.125 xRT 184 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.37 wall 0.123 xRT 185 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 186 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 187 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.07 CPU 0.368 xRT 188 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.99 wall 1.027 xRT 189 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.56 CPU 0.192 xRT 190 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.56 wall 0.193 xRT 191 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 192 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 193 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.93 CPU 0.355 xRT 194 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.69 wall 1.028 xRT 195 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.29 CPU 0.111 xRT 196 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.29 wall 0.109 xRT 197 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 198 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 199 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.47 CPU 0.332 xRT 200 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.91 wall 1.108 xRT 201 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.40 CPU 0.090 xRT 202 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.40 wall 0.090 xRT 203 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 204 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 205 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.55 CPU 0.348 xRT 206 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.77 wall 1.121 xRT 207 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.165 xRT 208 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.26 wall 0.164 xRT 209 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 210 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 211 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.42 CPU 0.350 xRT 212 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.26 wall 1.050 xRT 213 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.183 xRT 214 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.21 wall 0.179 xRT 215 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 216 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 217 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.93 CPU 0.371 xRT 218 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.92 wall 1.164 xRT 219 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.187 xRT 220 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.185 xRT 221 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 222 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 223 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.11 CPU 0.344 xRT 224 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.31 wall 1.023 xRT 225 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.146 xRT 226 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.144 xRT 227 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 228 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 229 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.00 CPU 0.365 xRT 230 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.82 wall 1.030 xRT 231 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.168 xRT 232 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.47 wall 0.170 xRT 233 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 234 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 235 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.14 CPU 0.357 xRT 236 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.30 wall 1.036 xRT 237 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.51 CPU 0.160 xRT 238 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.52 wall 0.162 xRT 239 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 240 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 241 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.69 CPU 0.358 xRT 242 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.00 wall 1.034 xRT 243 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.32 CPU 0.166 xRT 244 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.165 xRT 245 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 246 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 247 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.99 CPU 0.350 xRT 248 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.89 wall 1.021 xRT 249 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.45 CPU 0.159 xRT 250 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.45 wall 0.158 xRT 251 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 252 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 253 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.67 CPU 0.364 xRT 254 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.90 wall 1.033 xRT 255 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.147 xRT 256 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.27 wall 0.148 xRT 257 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 258 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 259 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.73 CPU 0.358 xRT 260 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.039 xRT 261 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.33 CPU 0.162 xRT 262 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.159 xRT 263 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 264 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 265 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.94 CPU 0.363 xRT 266 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.85 wall 1.486 xRT 267 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.178 xRT 268 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.179 xRT 269 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 270 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 271 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.26 CPU 0.366 xRT 272 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.62 wall 1.342 xRT 273 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.41 CPU 0.119 xRT 274 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.41 wall 0.120 xRT 275 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 276 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 277 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.27 CPU 0.370 xRT 278 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.49 wall 1.019 xRT 279 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.40 CPU 0.117 xRT 280 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.41 wall 0.119 xRT 281 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 282 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 283 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.70 CPU 0.355 xRT 284 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.174 xRT 285 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.24 CPU 0.122 xRT 286 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.25 wall 0.126 xRT 287 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 288 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 289 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.10 CPU 0.361 xRT 290 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.15 wall 1.032 xRT 291 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.141 xRT 292 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.44 wall 0.143 xRT 293 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 294 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 295 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.96 CPU 0.361 xRT 296 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.73 wall 1.027 xRT 297 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.42 CPU 0.158 xRT 298 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.158 xRT 299 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 300 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 301 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.64 CPU 0.366 xRT 302 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.83 wall 1.046 xRT 303 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.23 CPU 0.131 xRT 304 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.23 wall 0.130 xRT 305 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 306 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 307 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.68 CPU 0.358 xRT 308 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.97 wall 1.034 xRT 309 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.29 CPU 0.153 xRT 310 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.28 wall 0.149 xRT 311 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 312 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 313 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.32 CPU 0.427 xRT 314 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.81 wall 1.080 xRT 315 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.17 CPU 0.227 xRT 316 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.16 wall 0.220 xRT 317 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 318 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.003 xRT 319 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.09 CPU 0.354 xRT 320 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.17 wall 1.031 xRT 321 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.149 xRT 322 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.148 xRT 323 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 324 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 325 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.47 CPU 0.379 xRT 326 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.81 wall 1.240 xRT 327 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.111 xRT 328 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.44 wall 0.112 xRT 329 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 330 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 331 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.94 CPU 0.363 xRT 332 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.67 wall 1.029 xRT 333 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.131 xRT 334 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.131 xRT 335 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 336 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 337 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.85 CPU 0.381 xRT 338 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.036 xRT 339 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.193 xRT 340 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.43 wall 0.193 xRT 341 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 342 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 343 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.04 CPU 0.395 xRT 344 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.72 wall 1.415 xRT 345 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.175 xRT 346 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.175 xRT 347 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 348 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 349 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.05 CPU 0.382 xRT 350 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.95 wall 1.072 xRT 351 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.54 CPU 0.196 xRT 352 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.54 wall 0.197 xRT 353 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 354 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.002 xRT 355 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.68 CPU 0.415 xRT 356 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.47 wall 1.508 xRT 357 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.36 CPU 0.220 xRT 358 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.222 xRT 359 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 360 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 361 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.87 CPU 0.363 xRT 362 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.02 wall 1.258 xRT 363 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.142 xRT 364 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.143 xRT 365 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 366 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 367 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.41 CPU 0.373 xRT 368 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.26 wall 1.142 xRT 369 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.13 CPU 0.118 xRT 370 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.13 wall 0.118 xRT 371 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 372 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 373 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.84 CPU 0.411 xRT 374 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.55 wall 1.016 xRT 375 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.08 CPU 0.241 xRT 376 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.08 wall 0.241 xRT 377 | INFO: ngram_search.c(874): bestpath 0.06 CPU 0.013 xRT 378 | INFO: ngram_search.c(877): bestpath 0.06 wall 0.013 xRT 379 | INFO: ngram_search_fwdtree.c(432): TOTAL fwdtree 53.54 CPU 0.368 xRT 380 | INFO: ngram_search_fwdtree.c(435): TOTAL fwdtree 167.59 wall 1.153 xRT 381 | INFO: ngram_search_fwdflat.c(176): TOTAL fwdflat 21.99 CPU 0.151 xRT 382 | INFO: ngram_search_fwdflat.c(179): TOTAL fwdflat 22.00 wall 0.151 xRT 383 | INFO: ngram_search.c(303): TOTAL bestpath 0.17 CPU 0.001 xRT 384 | INFO: ngram_search.c(306): TOTAL bestpath 0.18 wall 0.001 xRT 385 | -------------------------------------------------------------------------------- /recoMic/results_pi3/pi3_results.log: -------------------------------------------------------------------------------- 1 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.70 CPU 0.376 xRT 2 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.44 wall 1.314 xRT 3 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.253 xRT 4 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.47 wall 0.253 xRT 5 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 6 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.006 xRT 7 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.01 CPU 0.365 xRT 8 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.40 wall 1.229 xRT 9 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.30 CPU 0.108 xRT 10 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.30 wall 0.107 xRT 11 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 12 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 13 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.63 CPU 0.375 xRT 14 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.35 wall 1.398 xRT 15 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.36 CPU 0.214 xRT 16 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.217 xRT 17 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 18 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 19 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.78 CPU 0.379 xRT 20 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.86 wall 1.389 xRT 21 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.41 CPU 0.199 xRT 22 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.203 xRT 23 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 24 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 25 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.76 CPU 0.362 xRT 26 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.18 wall 1.039 xRT 27 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.162 xRT 28 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.163 xRT 29 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 30 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 31 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.48 CPU 0.375 xRT 32 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.39 wall 1.086 xRT 33 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.19 CPU 0.148 xRT 34 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.19 wall 0.145 xRT 35 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 36 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 37 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.08 CPU 0.361 xRT 38 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.05 wall 1.021 xRT 39 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.45 CPU 0.151 xRT 40 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.45 wall 0.149 xRT 41 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 42 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 43 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.75 CPU 0.366 xRT 44 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.034 xRT 45 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.132 xRT 46 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.27 wall 0.133 xRT 47 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 48 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 49 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.59 CPU 0.351 xRT 50 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.74 wall 1.035 xRT 51 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.155 xRT 52 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.25 wall 0.150 xRT 53 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 54 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 55 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.75 CPU 0.359 xRT 56 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.18 wall 1.045 xRT 57 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.105 xRT 58 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.22 wall 0.107 xRT 59 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 60 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 61 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.27 CPU 0.422 xRT 62 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.71 wall 1.114 xRT 63 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.13 CPU 0.203 xRT 64 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.13 wall 0.210 xRT 65 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 66 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.002 xRT 67 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.73 CPU 0.372 xRT 68 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.09 wall 1.067 xRT 69 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.32 CPU 0.163 xRT 70 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.163 xRT 71 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 72 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 73 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.66 CPU 0.361 xRT 74 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.90 wall 1.037 xRT 75 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.35 CPU 0.191 xRT 76 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.195 xRT 77 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 78 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 79 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.69 CPU 0.381 xRT 80 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.82 wall 1.559 xRT 81 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.31 CPU 0.171 xRT 82 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.31 wall 0.170 xRT 83 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 84 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 85 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.35 CPU 0.422 xRT 86 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.83 wall 2.204 xRT 87 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.10 CPU 0.120 xRT 88 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.10 wall 0.116 xRT 89 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 90 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 91 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.36 CPU 0.419 xRT 92 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.93 wall 1.085 xRT 93 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.15 CPU 0.174 xRT 94 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.15 wall 0.176 xRT 95 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.012 xRT 96 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 97 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.61 CPU 0.367 xRT 98 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.73 wall 1.044 xRT 99 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.18 CPU 0.108 xRT 100 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.18 wall 0.109 xRT 101 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 102 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 103 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.67 CPU 0.356 xRT 104 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.96 wall 1.041 xRT 105 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.117 xRT 106 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.22 wall 0.117 xRT 107 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 108 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 109 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.74 CPU 0.354 xRT 110 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.16 wall 1.032 xRT 111 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.19 CPU 0.091 xRT 112 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.19 wall 0.093 xRT 113 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 114 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 115 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.14 CPU 0.344 xRT 116 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.40 wall 1.027 xRT 117 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.142 xRT 118 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.140 xRT 119 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 120 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 121 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.54 CPU 0.380 xRT 122 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.495 xRT 123 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.24 CPU 0.169 xRT 124 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.23 wall 0.164 xRT 125 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 126 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 127 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.92 CPU 0.376 xRT 128 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.27 wall 1.334 xRT 129 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.31 CPU 0.127 xRT 130 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.31 wall 0.128 xRT 131 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 132 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 133 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.76 CPU 0.376 xRT 134 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.73 wall 1.351 xRT 135 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.134 xRT 136 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.28 wall 0.138 xRT 137 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 138 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 139 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.35 CPU 0.366 xRT 140 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.86 wall 1.589 xRT 141 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.37 CPU 0.100 xRT 142 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.37 wall 0.099 xRT 143 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 144 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 145 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.56 CPU 0.368 xRT 146 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.58 wall 1.039 xRT 147 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.21 CPU 0.138 xRT 148 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.20 wall 0.134 xRT 149 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 150 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 151 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.80 CPU 0.374 xRT 152 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.079 xRT 153 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.38 CPU 0.178 xRT 154 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.38 wall 0.176 xRT 155 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 156 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.003 xRT 157 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.62 CPU 0.371 xRT 158 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.16 wall 1.291 xRT 159 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.156 xRT 160 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.26 wall 0.157 xRT 161 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 162 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 163 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.84 CPU 0.368 xRT 164 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.35 wall 1.029 xRT 165 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.50 CPU 0.219 xRT 166 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.50 wall 0.218 xRT 167 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 168 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.003 xRT 169 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.98 CPU 0.344 xRT 170 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.99 wall 1.048 xRT 171 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.42 CPU 0.147 xRT 172 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.146 xRT 173 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 174 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 175 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.48 CPU 0.360 xRT 176 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.17 wall 1.015 xRT 177 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.33 CPU 0.080 xRT 178 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.083 xRT 179 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.002 xRT 180 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 181 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.13 CPU 0.370 xRT 182 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.23 wall 1.387 xRT 183 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.38 CPU 0.125 xRT 184 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.37 wall 0.123 xRT 185 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 186 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 187 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.07 CPU 0.368 xRT 188 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.99 wall 1.027 xRT 189 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.56 CPU 0.192 xRT 190 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.56 wall 0.193 xRT 191 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 192 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 193 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.93 CPU 0.355 xRT 194 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.69 wall 1.028 xRT 195 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.29 CPU 0.111 xRT 196 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.29 wall 0.109 xRT 197 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 198 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 199 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.47 CPU 0.332 xRT 200 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.91 wall 1.108 xRT 201 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.40 CPU 0.090 xRT 202 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.40 wall 0.090 xRT 203 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 204 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 205 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.55 CPU 0.348 xRT 206 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.77 wall 1.121 xRT 207 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.26 CPU 0.165 xRT 208 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.26 wall 0.164 xRT 209 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 210 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 211 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.42 CPU 0.350 xRT 212 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.26 wall 1.050 xRT 213 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.22 CPU 0.183 xRT 214 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.21 wall 0.179 xRT 215 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 216 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 217 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.93 CPU 0.371 xRT 218 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.92 wall 1.164 xRT 219 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.187 xRT 220 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.185 xRT 221 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 222 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 223 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.11 CPU 0.344 xRT 224 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.31 wall 1.023 xRT 225 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.47 CPU 0.146 xRT 226 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.144 xRT 227 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 228 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 229 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.00 CPU 0.365 xRT 230 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.82 wall 1.030 xRT 231 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.168 xRT 232 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.47 wall 0.170 xRT 233 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 234 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 235 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.14 CPU 0.357 xRT 236 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.30 wall 1.036 xRT 237 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.51 CPU 0.160 xRT 238 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.52 wall 0.162 xRT 239 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 240 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 241 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.69 CPU 0.358 xRT 242 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.00 wall 1.034 xRT 243 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.32 CPU 0.166 xRT 244 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.165 xRT 245 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.005 xRT 246 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 247 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.99 CPU 0.350 xRT 248 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.89 wall 1.021 xRT 249 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.45 CPU 0.159 xRT 250 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.45 wall 0.158 xRT 251 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 252 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 253 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.67 CPU 0.364 xRT 254 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.90 wall 1.033 xRT 255 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.27 CPU 0.147 xRT 256 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.27 wall 0.148 xRT 257 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 258 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 259 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.73 CPU 0.358 xRT 260 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.12 wall 1.039 xRT 261 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.33 CPU 0.162 xRT 262 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.32 wall 0.159 xRT 263 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 264 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 265 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.94 CPU 0.363 xRT 266 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.85 wall 1.486 xRT 267 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.178 xRT 268 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.179 xRT 269 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 270 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 271 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.26 CPU 0.366 xRT 272 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.62 wall 1.342 xRT 273 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.41 CPU 0.119 xRT 274 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.41 wall 0.120 xRT 275 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 276 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 277 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.27 CPU 0.370 xRT 278 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.49 wall 1.019 xRT 279 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.40 CPU 0.117 xRT 280 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.41 wall 0.119 xRT 281 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 282 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 283 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.70 CPU 0.355 xRT 284 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.174 xRT 285 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.24 CPU 0.122 xRT 286 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.25 wall 0.126 xRT 287 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 288 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 289 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.10 CPU 0.361 xRT 290 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.15 wall 1.032 xRT 291 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.141 xRT 292 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.44 wall 0.143 xRT 293 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 294 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 295 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.96 CPU 0.361 xRT 296 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.73 wall 1.027 xRT 297 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.42 CPU 0.158 xRT 298 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.42 wall 0.158 xRT 299 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 300 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 301 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.64 CPU 0.366 xRT 302 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.83 wall 1.046 xRT 303 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.23 CPU 0.131 xRT 304 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.23 wall 0.130 xRT 305 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 306 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 307 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.68 CPU 0.358 xRT 308 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.97 wall 1.034 xRT 309 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.29 CPU 0.153 xRT 310 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.28 wall 0.149 xRT 311 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 312 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 313 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.32 CPU 0.427 xRT 314 | INFO: ngram_search_fwdtree.c(1570): fwdtree 0.81 wall 1.080 xRT 315 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.17 CPU 0.227 xRT 316 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.16 wall 0.220 xRT 317 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 318 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.003 xRT 319 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.09 CPU 0.354 xRT 320 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.17 wall 1.031 xRT 321 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.149 xRT 322 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.148 xRT 323 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 324 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 325 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.47 CPU 0.379 xRT 326 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.81 wall 1.240 xRT 327 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.111 xRT 328 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.44 wall 0.112 xRT 329 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.003 xRT 330 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 331 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.94 CPU 0.363 xRT 332 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.67 wall 1.029 xRT 333 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.131 xRT 334 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.131 xRT 335 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 336 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 337 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.85 CPU 0.381 xRT 338 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.31 wall 1.036 xRT 339 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.43 CPU 0.193 xRT 340 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.43 wall 0.193 xRT 341 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 342 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 343 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.04 CPU 0.395 xRT 344 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.72 wall 1.415 xRT 345 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.46 CPU 0.175 xRT 346 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.46 wall 0.175 xRT 347 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 348 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 349 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.05 CPU 0.382 xRT 350 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.95 wall 1.072 xRT 351 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.54 CPU 0.196 xRT 352 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.54 wall 0.197 xRT 353 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 354 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.002 xRT 355 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.68 CPU 0.415 xRT 356 | INFO: ngram_search_fwdtree.c(1570): fwdtree 2.47 wall 1.508 xRT 357 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.36 CPU 0.220 xRT 358 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.36 wall 0.222 xRT 359 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 360 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 361 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.87 CPU 0.363 xRT 362 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.02 wall 1.258 xRT 363 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.34 CPU 0.142 xRT 364 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.34 wall 0.143 xRT 365 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 366 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.001 xRT 367 | INFO: ngram_search_fwdtree.c(1567): fwdtree 0.41 CPU 0.373 xRT 368 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.26 wall 1.142 xRT 369 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.13 CPU 0.118 xRT 370 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.13 wall 0.118 xRT 371 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 372 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.000 xRT 373 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.84 CPU 0.411 xRT 374 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.55 wall 1.016 xRT 375 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.08 CPU 0.241 xRT 376 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.08 wall 0.241 xRT 377 | INFO: ngram_search.c(874): bestpath 0.06 CPU 0.013 xRT 378 | INFO: ngram_search.c(877): bestpath 0.06 wall 0.013 xRT 379 | INFO: ngram_search_fwdtree.c(432): TOTAL fwdtree 53.54 CPU 0.368 xRT 380 | INFO: ngram_search_fwdtree.c(435): TOTAL fwdtree 167.59 wall 1.153 xRT 381 | INFO: ngram_search_fwdflat.c(176): TOTAL fwdflat 21.99 CPU 0.151 xRT 382 | INFO: ngram_search_fwdflat.c(179): TOTAL fwdflat 22.00 wall 0.151 xRT 383 | INFO: ngram_search.c(303): TOTAL bestpath 0.17 CPU 0.001 xRT 384 | INFO: ngram_search.c(306): TOTAL bestpath 0.18 wall 0.001 xRT 385 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/psphinx_10LM_PiBplus.log: -------------------------------------------------------------------------------- 1 | INFO: cmd_ln.c(697): Parsing command line: 2 | \ 3 | -lowerf 130 \ 4 | -upperf 6800 \ 5 | -nfilt 25 \ 6 | -transform dct \ 7 | -lifter 22 \ 8 | -feat 1s_c_d_dd \ 9 | -svspec 0-12/13-25/26-38 \ 10 | -agc none \ 11 | -cmn current \ 12 | -varnorm no \ 13 | -model ptm \ 14 | -cmninit 40,3,-1 15 | 16 | Current configuration: 17 | [NAME] [DEFLT] [VALUE] 18 | -agc none none 19 | -agcthresh 2.0 2.000000e+00 20 | -alpha 0.97 9.700000e-01 21 | -ceplen 13 13 22 | -cmn current current 23 | -cmninit 8.0 40,3,-1 24 | -dither no no 25 | -doublebw no no 26 | -feat 1s_c_d_dd 1s_c_d_dd 27 | -frate 100 100 28 | -input_endian little little 29 | -lda 30 | -ldadim 0 0 31 | -lifter 0 22 32 | -logspec no no 33 | -lowerf 133.33334 1.300000e+02 34 | -ncep 13 13 35 | -nfft 512 512 36 | -nfilt 40 25 37 | -remove_dc no no 38 | -remove_noise yes yes 39 | -remove_silence yes yes 40 | -round_filters yes yes 41 | -samprate 16000 1.600000e+04 42 | -seed -1 -1 43 | -smoothspec no no 44 | -svspec 0-12/13-25/26-38 45 | -transform legacy dct 46 | -unit_area yes yes 47 | -upperf 6855.4976 6.800000e+03 48 | -vad_postspeech 50 50 49 | -vad_prespeech 10 10 50 | -vad_threshold 2.0 2.000000e+00 51 | -varnorm no no 52 | -verbose no yes 53 | -warp_params 54 | -warp_type inverse_linear inverse_linear 55 | -wlen 0.025625 2.562500e-02 56 | 57 | INFO: acmod.c(252): Parsed model-specific feature parameters from cmusphinx-5prealpha-en-us-ptm-2.0//feat.params 58 | INFO: fe_interface.c(177): Current FE Parameters: 59 | INFO: fe_interface.c(178): Sampling Rate: 16000.000000 60 | INFO: fe_interface.c(179): Frame Size: 410 61 | INFO: fe_interface.c(180): Frame Shift: 160 62 | INFO: fe_interface.c(181): FFT Size: 512 63 | INFO: fe_interface.c(183): Lower Frequency: 130 64 | INFO: fe_interface.c(185): Upper Frequency: 6800 65 | INFO: fe_interface.c(186): Number of filters: 25 66 | INFO: fe_interface.c(187): Number of Overflow Samps: 0 67 | INFO: fe_interface.c(188): Start Utt Status: 0 68 | INFO: fe_interface.c(190): Will not remove DC offset at frame level 69 | INFO: fe_interface.c(196): Will not add dither to audio 70 | INFO: fe_interface.c(200): Will apply sine-curve liftering, period 22 71 | INFO: fe_interface.c(203): Will normalize filters to unit area 72 | INFO: fe_interface.c(205): Will round filter frequencies to DFT points 73 | INFO: fe_interface.c(207): Will not use double bandwidth in mel filter 74 | INFO: feat.c(715): Initializing feature stream to type: '1s_c_d_dd', ceplen=13, CMN='current', VARNORM='no', AGC='none' 75 | INFO: cmn.c(143): mean[0]= 12.00, mean[1..12]= 0.0 76 | INFO: acmod.c(171): Using subvector specification 0-12/13-25/26-38 77 | INFO: mdef.c(518): Reading model definition: cmusphinx-5prealpha-en-us-ptm-2.0//mdef 78 | INFO: bin_mdef.c(181): Allocating 142108 * 8 bytes (1110 KiB) for CD tree 79 | INFO: tmat.c(206): Reading HMM transition probability matrices: cmusphinx-5prealpha-en-us-ptm-2.0//transition_matrices 80 | INFO: acmod.c(124): Attempting to use PTM computation module 81 | INFO: ms_gauden.c(198): Reading mixture gaussian parameter: cmusphinx-5prealpha-en-us-ptm-2.0//means 82 | INFO: ms_gauden.c(292): 42 codebook, 3 feature, size: 83 | INFO: ms_gauden.c(294): 128x13 84 | INFO: ms_gauden.c(294): 128x13 85 | INFO: ms_gauden.c(294): 128x13 86 | INFO: ms_gauden.c(198): Reading mixture gaussian parameter: cmusphinx-5prealpha-en-us-ptm-2.0//variances 87 | INFO: ms_gauden.c(292): 42 codebook, 3 feature, size: 88 | INFO: ms_gauden.c(294): 128x13 89 | INFO: ms_gauden.c(294): 128x13 90 | INFO: ms_gauden.c(294): 128x13 91 | INFO: ms_gauden.c(354): 222 variance values floored 92 | INFO: ptm_mgau.c(476): Loading senones from dump file cmusphinx-5prealpha-en-us-ptm-2.0//sendump 93 | INFO: ptm_mgau.c(500): BEGIN FILE FORMAT DESCRIPTION 94 | INFO: ptm_mgau.c(563): Rows: 128, Columns: 5126 95 | INFO: ptm_mgau.c(595): Using memory-mapped I/O for senones 96 | INFO: ptm_mgau.c(835): Maximum top-N: 4 97 | INFO: phone_loop_search.c(115): State beam -225 Phone exit beam -225 Insertion penalty 0 98 | INFO: dict.c(320): Allocating 4285 * 20 bytes (83 KiB) for word entries 99 | INFO: dict.c(333): Reading main dictionary: dictionary.dic 100 | ERROR: "dict.c", line 205: Line 58: Failed to add the word 'GO' (duplicate?); ignored 101 | ERROR: "dict.c", line 205: Line 76: Failed to add the word 'KITCHEN' (duplicate?); ignored 102 | ERROR: "dict.c", line 205: Line 98: Failed to add the word 'ON' (duplicate?); ignored 103 | ERROR: "dict.c", line 205: Line 99: Failed to add the word 'ON(2)' (duplicate?); ignored 104 | ERROR: "dict.c", line 205: Line 145: Failed to add the word 'THE' (duplicate?); ignored 105 | ERROR: "dict.c", line 205: Line 146: Failed to add the word 'THE(2)' (duplicate?); ignored 106 | ERROR: "dict.c", line 205: Line 155: Failed to add the word 'TURN' (duplicate?); ignored 107 | INFO: dict.c(213): Allocated 0 KiB for strings, 1 KiB for phones 108 | INFO: dict.c(336): 177 words read 109 | INFO: dict.c(342): Reading filler dictionary: cmusphinx-5prealpha-en-us-ptm-2.0//noisedict 110 | INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones 111 | INFO: dict.c(345): 5 words read 112 | INFO: dict2pid.c(396): Building PID tables for dictionary 113 | INFO: dict2pid.c(406): Allocating 42^3 * 2 bytes (144 KiB) for word-initial triphones 114 | INFO: dict2pid.c(132): Allocated 21336 bytes (20 KiB) for word-final triphones 115 | INFO: dict2pid.c(196): Allocated 21336 bytes (20 KiB) for single-phone word triphones 116 | INFO: ngram_model_arpa.c(477): ngrams 1=136, 2=278, 3=295 117 | INFO: ngram_model_arpa.c(135): Reading unigrams 118 | INFO: ngram_model_arpa.c(516): 136 = #unigrams created 119 | INFO: ngram_model_arpa.c(195): Reading bigrams 120 | INFO: ngram_model_arpa.c(534): 278 = #bigrams created 121 | INFO: ngram_model_arpa.c(535): 36 = #prob2 entries 122 | INFO: ngram_model_arpa.c(543): 28 = #bo_wt2 entries 123 | INFO: ngram_model_arpa.c(292): Reading trigrams 124 | INFO: ngram_model_arpa.c(556): 295 = #trigrams created 125 | INFO: ngram_model_arpa.c(557): 21 = #prob3 entries 126 | INFO: ngram_search_fwdtree.c(99): 112 unique initial diphones 127 | INFO: ngram_search_fwdtree.c(148): 0 root, 0 non-root channels, 9 single-phone words 128 | INFO: ngram_search_fwdtree.c(186): Creating search tree 129 | INFO: ngram_search_fwdtree.c(192): before: 0 root, 0 non-root channels, 9 single-phone words 130 | INFO: ngram_search_fwdtree.c(326): after: max nonroot chan increased to 438 131 | INFO: ngram_search_fwdtree.c(339): after: 106 root, 310 non-root channels, 8 single-phone words 132 | INFO: ngram_search_fwdflat.c(157): fwdflat: min_ef_width = 4, max_sf_win = 25 133 | INFO: cmn_prior.c(131): cmn_prior_update: from < 40.00 3.00 -1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 > 134 | INFO: cmn_prior.c(149): cmn_prior_update: to < 39.36 0.68 -9.21 -1.18 -8.38 -1.43 3.41 0.58 -1.36 6.82 4.03 4.75 -0.84 > 135 | INFO: ngram_search_fwdtree.c(1553): 1373 words recognized (14/fr) 136 | INFO: ngram_search_fwdtree.c(1555): 86270 senones evaluated (854/fr) 137 | INFO: ngram_search_fwdtree.c(1559): 57327 channels searched (567/fr), 10282 1st, 33009 last 138 | INFO: ngram_search_fwdtree.c(1562): 2105 words for which last channels evaluated (20/fr) 139 | INFO: ngram_search_fwdtree.c(1564): 2266 candidate words for entering last phone (22/fr) 140 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.86 CPU 1.842 xRT 141 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.81 wall 3.770 xRT 142 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 48 words 143 | INFO: ngram_search_fwdflat.c(945): 1111 words recognized (11/fr) 144 | INFO: ngram_search_fwdflat.c(947): 60628 senones evaluated (600/fr) 145 | INFO: ngram_search_fwdflat.c(949): 64652 channels searched (640/fr) 146 | INFO: ngram_search_fwdflat.c(951): 3598 words searched (35/fr) 147 | INFO: ngram_search_fwdflat.c(954): 1942 word transitions (19/fr) 148 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.04 CPU 1.030 xRT 149 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.09 wall 1.074 xRT 150 | INFO: ngram_search.c(1252): lattice start node .0 end node .52 151 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 152 | INFO: ngram_search.c(1383): Lattice has 178 nodes, 559 links 153 | INFO: ps_lattice.c(1380): Bestpath score: -2468 154 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:52:99) = -121375 155 | INFO: ps_lattice.c(1441): Joint P(O,S) = -167753 P(S|O) = -46378 156 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.010 xRT 157 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.015 xRT 158 | INFO: cmn_prior.c(131): cmn_prior_update: from < 39.36 0.68 -9.21 -1.18 -8.38 -1.43 3.41 0.58 -1.36 6.82 4.03 4.75 -0.84 > 159 | INFO: cmn_prior.c(149): cmn_prior_update: to < 40.86 -1.41 -5.44 -2.05 -4.29 0.24 1.25 -2.11 -0.61 4.50 4.29 1.71 -0.98 > 160 | INFO: ngram_search_fwdtree.c(1553): 1009 words recognized (11/fr) 161 | INFO: ngram_search_fwdtree.c(1555): 84091 senones evaluated (934/fr) 162 | INFO: ngram_search_fwdtree.c(1559): 51382 channels searched (570/fr), 9116 1st, 27624 last 163 | INFO: ngram_search_fwdtree.c(1562): 1763 words for which last channels evaluated (19/fr) 164 | INFO: ngram_search_fwdtree.c(1564): 2965 candidate words for entering last phone (32/fr) 165 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.69 CPU 1.878 xRT 166 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.54 wall 3.938 xRT 167 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 45 words 168 | INFO: ngram_search_fwdflat.c(945): 768 words recognized (9/fr) 169 | INFO: ngram_search_fwdflat.c(947): 68479 senones evaluated (761/fr) 170 | INFO: ngram_search_fwdflat.c(949): 67795 channels searched (753/fr) 171 | INFO: ngram_search_fwdflat.c(951): 3490 words searched (38/fr) 172 | INFO: ngram_search_fwdflat.c(954): 2210 word transitions (24/fr) 173 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.00 CPU 1.111 xRT 174 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.00 wall 1.108 xRT 175 | INFO: ngram_search.c(1252): lattice start node .0 end node .49 176 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 177 | INFO: ngram_search.c(1383): Lattice has 192 nodes, 285 links 178 | INFO: ps_lattice.c(1380): Bestpath score: -2589 179 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:49:88) = -163035 180 | INFO: ps_lattice.c(1441): Joint P(O,S) = -191338 P(S|O) = -28303 181 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 182 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.008 xRT 183 | INFO: cmn_prior.c(131): cmn_prior_update: from < 40.86 -1.41 -5.44 -2.05 -4.29 0.24 1.25 -2.11 -0.61 4.50 4.29 1.71 -0.98 > 184 | INFO: cmn_prior.c(149): cmn_prior_update: to < 45.34 0.46 -5.03 -1.42 -4.57 2.28 0.56 -2.87 -1.82 3.37 3.84 2.63 -1.02 > 185 | INFO: ngram_search_fwdtree.c(1553): 1009 words recognized (9/fr) 186 | INFO: ngram_search_fwdtree.c(1555): 92304 senones evaluated (782/fr) 187 | INFO: ngram_search_fwdtree.c(1559): 56822 channels searched (481/fr), 11224 1st, 30828 last 188 | INFO: ngram_search_fwdtree.c(1562): 1831 words for which last channels evaluated (15/fr) 189 | INFO: ngram_search_fwdtree.c(1564): 2518 candidate words for entering last phone (21/fr) 190 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.15 CPU 1.822 xRT 191 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.82 wall 3.237 xRT 192 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 38 words 193 | INFO: ngram_search_fwdflat.c(945): 519 words recognized (4/fr) 194 | INFO: ngram_search_fwdflat.c(947): 65288 senones evaluated (553/fr) 195 | INFO: ngram_search_fwdflat.c(949): 57429 channels searched (486/fr) 196 | INFO: ngram_search_fwdflat.c(951): 3257 words searched (27/fr) 197 | INFO: ngram_search_fwdflat.c(954): 2209 word transitions (18/fr) 198 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.14 CPU 0.966 xRT 199 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.15 wall 0.974 xRT 200 | INFO: ngram_search.c(1252): lattice start node .0 end node .78 201 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 202 | INFO: ngram_search.c(1383): Lattice has 135 nodes, 129 links 203 | INFO: ps_lattice.c(1380): Bestpath score: -4833 204 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:78:116) = -298824 205 | INFO: ps_lattice.c(1441): Joint P(O,S) = -307012 P(S|O) = -8188 206 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.009 xRT 207 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.003 xRT 208 | INFO: cmn_prior.c(131): cmn_prior_update: from < 45.34 0.46 -5.03 -1.42 -4.57 2.28 0.56 -2.87 -1.82 3.37 3.84 2.63 -1.02 > 209 | INFO: cmn_prior.c(149): cmn_prior_update: to < 42.67 0.53 -6.37 -0.66 -3.75 2.97 1.71 -0.51 -0.73 3.04 5.21 2.05 -1.03 > 210 | INFO: ngram_search_fwdtree.c(1553): 2312 words recognized (12/fr) 211 | INFO: ngram_search_fwdtree.c(1555): 167405 senones evaluated (845/fr) 212 | INFO: ngram_search_fwdtree.c(1559): 99234 channels searched (501/fr), 20531 1st, 49703 last 213 | INFO: ngram_search_fwdtree.c(1562): 3533 words for which last channels evaluated (17/fr) 214 | INFO: ngram_search_fwdtree.c(1564): 5344 candidate words for entering last phone (26/fr) 215 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.57 CPU 1.803 xRT 216 | INFO: ngram_search_fwdtree.c(1570): fwdtree 6.02 wall 3.038 xRT 217 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 56 words 218 | INFO: ngram_search_fwdflat.c(945): 1868 words recognized (9/fr) 219 | INFO: ngram_search_fwdflat.c(947): 98260 senones evaluated (496/fr) 220 | INFO: ngram_search_fwdflat.c(949): 106392 channels searched (537/fr) 221 | INFO: ngram_search_fwdflat.c(951): 5604 words searched (28/fr) 222 | INFO: ngram_search_fwdflat.c(954): 3504 word transitions (17/fr) 223 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.83 CPU 0.924 xRT 224 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.85 wall 0.936 xRT 225 | INFO: ngram_search.c(1252): lattice start node .0 end node .122 226 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 227 | INFO: ngram_search.c(1383): Lattice has 300 nodes, 1348 links 228 | INFO: ps_lattice.c(1380): Bestpath score: -3585 229 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:122:196) = -277406 230 | INFO: ps_lattice.c(1441): Joint P(O,S) = -300795 P(S|O) = -23389 231 | INFO: ngram_search.c(874): bestpath 0.06 CPU 0.030 xRT 232 | INFO: ngram_search.c(877): bestpath 0.10 wall 0.050 xRT 233 | INFO: cmn_prior.c(131): cmn_prior_update: from < 42.67 0.53 -6.37 -0.66 -3.75 2.97 1.71 -0.51 -0.73 3.04 5.21 2.05 -1.03 > 234 | INFO: cmn_prior.c(149): cmn_prior_update: to < 44.87 1.11 -8.09 1.43 -2.72 3.45 1.11 -1.09 -1.08 2.93 4.31 1.85 -0.76 > 235 | INFO: ngram_search_fwdtree.c(1553): 1820 words recognized (14/fr) 236 | INFO: ngram_search_fwdtree.c(1555): 115048 senones evaluated (878/fr) 237 | INFO: ngram_search_fwdtree.c(1559): 77364 channels searched (590/fr), 12936 1st, 47497 last 238 | INFO: ngram_search_fwdtree.c(1562): 2927 words for which last channels evaluated (22/fr) 239 | INFO: ngram_search_fwdtree.c(1564): 2874 candidate words for entering last phone (21/fr) 240 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.60 CPU 1.985 xRT 241 | INFO: ngram_search_fwdtree.c(1570): fwdtree 6.49 wall 4.957 xRT 242 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 60 words 243 | INFO: ngram_search_fwdflat.c(945): 1640 words recognized (13/fr) 244 | INFO: ngram_search_fwdflat.c(947): 102617 senones evaluated (783/fr) 245 | INFO: ngram_search_fwdflat.c(949): 104182 channels searched (795/fr) 246 | INFO: ngram_search_fwdflat.c(951): 5336 words searched (40/fr) 247 | INFO: ngram_search_fwdflat.c(954): 3616 word transitions (27/fr) 248 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.59 CPU 1.214 xRT 249 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.64 wall 1.249 xRT 250 | INFO: ngram_search.c(1252): lattice start node .0 end node .90 251 | INFO: ngram_search.c(1278): Eliminated 0 nodes before end node 252 | INFO: ngram_search.c(1383): Lattice has 260 nodes, 1103 links 253 | INFO: ps_lattice.c(1380): Bestpath score: -6362 254 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:90:129) = -360548 255 | INFO: ps_lattice.c(1441): Joint P(O,S) = -427321 P(S|O) = -66773 256 | INFO: ngram_search.c(874): bestpath 0.05 CPU 0.038 xRT 257 | INFO: ngram_search.c(877): bestpath 0.04 wall 0.034 xRT 258 | INFO: cmn_prior.c(131): cmn_prior_update: from < 44.87 1.11 -8.09 1.43 -2.72 3.45 1.11 -1.09 -1.08 2.93 4.31 1.85 -0.76 > 259 | INFO: cmn_prior.c(149): cmn_prior_update: to < 45.64 1.58 -8.44 2.04 -2.96 3.79 0.85 -0.71 -1.28 2.37 4.14 1.16 -0.84 > 260 | INFO: ngram_search_fwdtree.c(1553): 1898 words recognized (17/fr) 261 | INFO: ngram_search_fwdtree.c(1555): 109689 senones evaluated (954/fr) 262 | INFO: ngram_search_fwdtree.c(1559): 79216 channels searched (688/fr), 11740 1st, 50748 last 263 | INFO: ngram_search_fwdtree.c(1562): 2906 words for which last channels evaluated (25/fr) 264 | INFO: ngram_search_fwdtree.c(1564): 3633 candidate words for entering last phone (31/fr) 265 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.22 CPU 1.930 xRT 266 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.55 wall 3.954 xRT 267 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 59 words 268 | INFO: ngram_search_fwdflat.c(945): 1310 words recognized (11/fr) 269 | INFO: ngram_search_fwdflat.c(947): 96123 senones evaluated (836/fr) 270 | INFO: ngram_search_fwdflat.c(949): 101289 channels searched (880/fr) 271 | INFO: ngram_search_fwdflat.c(951): 5127 words searched (44/fr) 272 | INFO: ngram_search_fwdflat.c(954): 3427 word transitions (29/fr) 273 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.44 CPU 1.252 xRT 274 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.56 wall 1.359 xRT 275 | INFO: ngram_search.c(1252): lattice start node .0 end node .73 276 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 277 | INFO: ngram_search.c(1383): Lattice has 263 nodes, 823 links 278 | INFO: ps_lattice.c(1380): Bestpath score: -5097 279 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:73:113) = -290307 280 | INFO: ps_lattice.c(1441): Joint P(O,S) = -302785 P(S|O) = -12478 281 | INFO: ngram_search.c(874): bestpath 0.02 CPU 0.018 xRT 282 | INFO: ngram_search.c(877): bestpath 0.02 wall 0.021 xRT 283 | INFO: cmn_prior.c(99): cmn_prior_update: from < 45.64 1.58 -8.44 2.04 -2.96 3.79 0.85 -0.71 -1.28 2.37 4.14 1.16 -0.84 > 284 | INFO: cmn_prior.c(116): cmn_prior_update: to < 46.63 1.67 -8.39 3.46 -3.35 4.26 0.46 -0.78 -1.83 1.35 3.99 0.80 -0.34 > 285 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.63 1.67 -8.39 3.46 -3.35 4.26 0.46 -0.78 -1.83 1.35 3.99 0.80 -0.34 > 286 | INFO: cmn_prior.c(149): cmn_prior_update: to < 46.50 1.20 -8.75 3.95 -3.44 3.96 0.29 -0.35 -1.57 1.34 4.55 0.99 -0.31 > 287 | INFO: ngram_search_fwdtree.c(1553): 1102 words recognized (11/fr) 288 | INFO: ngram_search_fwdtree.c(1555): 79520 senones evaluated (787/fr) 289 | INFO: ngram_search_fwdtree.c(1559): 49299 channels searched (488/fr), 10088 1st, 28315 last 290 | INFO: ngram_search_fwdtree.c(1562): 1779 words for which last channels evaluated (17/fr) 291 | INFO: ngram_search_fwdtree.c(1564): 1645 candidate words for entering last phone (16/fr) 292 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.84 CPU 1.822 xRT 293 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.41 wall 5.352 xRT 294 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 40 words 295 | INFO: ngram_search_fwdflat.c(945): 904 words recognized (9/fr) 296 | INFO: ngram_search_fwdflat.c(947): 59454 senones evaluated (589/fr) 297 | INFO: ngram_search_fwdflat.c(949): 55045 channels searched (545/fr) 298 | INFO: ngram_search_fwdflat.c(951): 3034 words searched (30/fr) 299 | INFO: ngram_search_fwdflat.c(954): 2072 word transitions (20/fr) 300 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.06 CPU 1.050 xRT 301 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.79 wall 1.774 xRT 302 | INFO: ngram_search.c(1252): lattice start node .0 end node .54 303 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 304 | INFO: ngram_search.c(1383): Lattice has 155 nodes, 199 links 305 | INFO: ps_lattice.c(1380): Bestpath score: -3731 306 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:54:99) = -252792 307 | INFO: ps_lattice.c(1441): Joint P(O,S) = -275918 P(S|O) = -23126 308 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 309 | INFO: ngram_search.c(877): bestpath 0.02 wall 0.016 xRT 310 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.50 1.20 -8.75 3.95 -3.44 3.96 0.29 -0.35 -1.57 1.34 4.55 0.99 -0.31 > 311 | INFO: cmn_prior.c(149): cmn_prior_update: to < 46.67 -2.54 -7.57 3.19 -3.04 3.47 0.02 -0.70 -1.80 0.55 4.16 0.57 -0.10 > 312 | INFO: ngram_search_fwdtree.c(1553): 1712 words recognized (12/fr) 313 | INFO: ngram_search_fwdtree.c(1555): 84687 senones evaluated (614/fr) 314 | INFO: ngram_search_fwdtree.c(1559): 56848 channels searched (411/fr), 9519 1st, 37348 last 315 | INFO: ngram_search_fwdtree.c(1562): 2418 words for which last channels evaluated (17/fr) 316 | INFO: ngram_search_fwdtree.c(1564): 1423 candidate words for entering last phone (10/fr) 317 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.91 CPU 2.109 xRT 318 | INFO: ngram_search_fwdtree.c(1570): fwdtree 6.65 wall 4.819 xRT 319 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 48 words 320 | INFO: ngram_search_fwdflat.c(945): 1081 words recognized (8/fr) 321 | INFO: ngram_search_fwdflat.c(947): 95273 senones evaluated (690/fr) 322 | INFO: ngram_search_fwdflat.c(949): 92074 channels searched (667/fr) 323 | INFO: ngram_search_fwdflat.c(951): 4699 words searched (34/fr) 324 | INFO: ngram_search_fwdflat.c(954): 2846 word transitions (20/fr) 325 | INFO: ngram_search_fwdflat.c(957): fwdflat 2.44 CPU 1.768 xRT 326 | INFO: ngram_search_fwdflat.c(960): fwdflat 5.77 wall 4.178 xRT 327 | INFO: ngram_search.c(1199): not found in last frame, using FACE.126 instead 328 | INFO: ngram_search.c(1252): lattice start node .0 end node FACE.3 329 | INFO: ngram_search.c(1278): Eliminated 122 nodes before end node 330 | INFO: ngram_search.c(1383): Lattice has 171 nodes, 1 links 331 | INFO: ps_lattice.c(1380): Bestpath score: -819 332 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(FACE:3:126) = -344674 333 | INFO: ps_lattice.c(1441): Joint P(O,S) = -344674 P(S|O) = 0 334 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 335 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.002 xRT 336 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.67 -2.54 -7.57 3.19 -3.04 3.47 0.02 -0.70 -1.80 0.55 4.16 0.57 -0.10 > 337 | INFO: cmn_prior.c(149): cmn_prior_update: to < 47.29 -4.06 -7.89 1.94 -3.08 3.24 -0.01 -0.60 -1.21 0.43 3.93 0.52 -0.49 > 338 | INFO: ngram_search_fwdtree.c(1553): 732 words recognized (10/fr) 339 | INFO: ngram_search_fwdtree.c(1555): 52813 senones evaluated (714/fr) 340 | INFO: ngram_search_fwdtree.c(1559): 30230 channels searched (408/fr), 6454 1st, 16888 last 341 | INFO: ngram_search_fwdtree.c(1562): 1215 words for which last channels evaluated (16/fr) 342 | INFO: ngram_search_fwdtree.c(1564): 961 candidate words for entering last phone (12/fr) 343 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.56 CPU 2.108 xRT 344 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.30 wall 5.812 xRT 345 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 23 words 346 | INFO: ngram_search_fwdflat.c(945): 546 words recognized (7/fr) 347 | INFO: ngram_search_fwdflat.c(947): 33010 senones evaluated (446/fr) 348 | INFO: ngram_search_fwdflat.c(949): 26389 channels searched (356/fr) 349 | INFO: ngram_search_fwdflat.c(951): 1587 words searched (21/fr) 350 | INFO: ngram_search_fwdflat.c(954): 776 word transitions (10/fr) 351 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.79 CPU 1.068 xRT 352 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.69 wall 2.279 xRT 353 | INFO: ngram_search.c(1252): lattice start node .0 end node .69 354 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 355 | INFO: ngram_search.c(1383): Lattice has 130 nodes, 639 links 356 | INFO: ps_lattice.c(1380): Bestpath score: -3238 357 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:69:72) = -193011 358 | INFO: ps_lattice.c(1441): Joint P(O,S) = -205478 P(S|O) = -12467 359 | INFO: ngram_search.c(874): bestpath 0.02 CPU 0.027 xRT 360 | INFO: ngram_search.c(877): bestpath 0.04 wall 0.053 xRT 361 | INFO: cmn_prior.c(99): cmn_prior_update: from < 47.29 -4.06 -7.89 1.94 -3.08 3.24 -0.01 -0.60 -1.21 0.43 3.93 0.52 -0.49 > 362 | INFO: cmn_prior.c(116): cmn_prior_update: to < 47.66 -4.59 -7.72 1.92 -3.17 3.50 -0.29 -0.56 -1.46 0.37 4.01 0.36 -0.53 > 363 | INFO: cmn_prior.c(131): cmn_prior_update: from < 47.66 -4.59 -7.72 1.92 -3.17 3.50 -0.29 -0.56 -1.46 0.37 4.01 0.36 -0.53 > 364 | INFO: cmn_prior.c(149): cmn_prior_update: to < 50.14 -4.43 -10.27 3.41 -3.97 2.87 0.29 -1.20 -1.37 1.26 5.89 0.43 -0.31 > 365 | INFO: ngram_search_fwdtree.c(1553): 1583 words recognized (10/fr) 366 | INFO: ngram_search_fwdtree.c(1555): 116909 senones evaluated (704/fr) 367 | INFO: ngram_search_fwdtree.c(1559): 72568 channels searched (437/fr), 15583 1st, 41048 last 368 | INFO: ngram_search_fwdtree.c(1562): 2631 words for which last channels evaluated (15/fr) 369 | INFO: ngram_search_fwdtree.c(1564): 2365 candidate words for entering last phone (14/fr) 370 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.03 CPU 1.825 xRT 371 | INFO: ngram_search_fwdtree.c(1570): fwdtree 6.32 wall 3.805 xRT 372 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 57 words 373 | INFO: ngram_search_fwdflat.c(945): 812 words recognized (5/fr) 374 | INFO: ngram_search_fwdflat.c(947): 88133 senones evaluated (531/fr) 375 | INFO: ngram_search_fwdflat.c(949): 83349 channels searched (502/fr) 376 | INFO: ngram_search_fwdflat.c(951): 4937 words searched (29/fr) 377 | INFO: ngram_search_fwdflat.c(954): 3642 word transitions (21/fr) 378 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.64 CPU 0.988 xRT 379 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.68 wall 1.012 xRT 380 | INFO: ngram_search.c(1252): lattice start node .0 end node .118 381 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 382 | INFO: ngram_search.c(1383): Lattice has 163 nodes, 100 links 383 | INFO: ps_lattice.c(1380): Bestpath score: -6342 384 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:118:164) = -368863 385 | INFO: ps_lattice.c(1441): Joint P(O,S) = -380503 P(S|O) = -11640 386 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.006 xRT 387 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.002 xRT 388 | INFO: ngram_search_fwdtree.c(432): TOTAL fwdtree 23.43 CPU 1.917 xRT 389 | INFO: ngram_search_fwdtree.c(435): TOTAL fwdtree 50.90 wall 4.165 xRT 390 | INFO: ngram_search_fwdflat.c(176): TOTAL fwdflat 13.97 CPU 1.143 xRT 391 | INFO: ngram_search_fwdflat.c(179): TOTAL fwdflat 19.21 wall 1.572 xRT 392 | INFO: ngram_search.c(303): TOTAL bestpath 0.18 CPU 0.015 xRT 393 | INFO: ngram_search.c(306): TOTAL bestpath 0.25 wall 0.021 xRT 394 | -------------------------------------------------------------------------------- /recoMic/results_piBplus/psphinx_14.log: -------------------------------------------------------------------------------- 1 | INFO: cmd_ln.c(697): Parsing command line: 2 | \ 3 | -lowerf 130 \ 4 | -upperf 6800 \ 5 | -nfilt 25 \ 6 | -transform dct \ 7 | -lifter 22 \ 8 | -feat 1s_c_d_dd \ 9 | -svspec 0-12/13-25/26-38 \ 10 | -agc none \ 11 | -cmn current \ 12 | -varnorm no \ 13 | -model ptm \ 14 | -cmninit 40,3,-1 15 | 16 | Current configuration: 17 | [NAME] [DEFLT] [VALUE] 18 | -agc none none 19 | -agcthresh 2.0 2.000000e+00 20 | -alpha 0.97 9.700000e-01 21 | -ceplen 13 13 22 | -cmn current current 23 | -cmninit 8.0 40,3,-1 24 | -dither no no 25 | -doublebw no no 26 | -feat 1s_c_d_dd 1s_c_d_dd 27 | -frate 100 100 28 | -input_endian little little 29 | -lda 30 | -ldadim 0 0 31 | -lifter 0 22 32 | -logspec no no 33 | -lowerf 133.33334 1.300000e+02 34 | -ncep 13 13 35 | -nfft 512 512 36 | -nfilt 40 25 37 | -remove_dc no no 38 | -remove_noise yes yes 39 | -remove_silence yes yes 40 | -round_filters yes yes 41 | -samprate 16000 1.600000e+04 42 | -seed -1 -1 43 | -smoothspec no no 44 | -svspec 0-12/13-25/26-38 45 | -transform legacy dct 46 | -unit_area yes yes 47 | -upperf 6855.4976 6.800000e+03 48 | -vad_postspeech 50 50 49 | -vad_prespeech 10 10 50 | -vad_threshold 2.0 2.000000e+00 51 | -varnorm no no 52 | -verbose no yes 53 | -warp_params 54 | -warp_type inverse_linear inverse_linear 55 | -wlen 0.025625 2.562500e-02 56 | 57 | INFO: acmod.c(252): Parsed model-specific feature parameters from cmusphinx-5prealpha-en-us-ptm-2.0//feat.params 58 | INFO: fe_interface.c(177): Current FE Parameters: 59 | INFO: fe_interface.c(178): Sampling Rate: 16000.000000 60 | INFO: fe_interface.c(179): Frame Size: 410 61 | INFO: fe_interface.c(180): Frame Shift: 160 62 | INFO: fe_interface.c(181): FFT Size: 512 63 | INFO: fe_interface.c(183): Lower Frequency: 130 64 | INFO: fe_interface.c(185): Upper Frequency: 6800 65 | INFO: fe_interface.c(186): Number of filters: 25 66 | INFO: fe_interface.c(187): Number of Overflow Samps: 0 67 | INFO: fe_interface.c(188): Start Utt Status: 0 68 | INFO: fe_interface.c(190): Will not remove DC offset at frame level 69 | INFO: fe_interface.c(196): Will not add dither to audio 70 | INFO: fe_interface.c(200): Will apply sine-curve liftering, period 22 71 | INFO: fe_interface.c(203): Will normalize filters to unit area 72 | INFO: fe_interface.c(205): Will round filter frequencies to DFT points 73 | INFO: fe_interface.c(207): Will not use double bandwidth in mel filter 74 | INFO: feat.c(715): Initializing feature stream to type: '1s_c_d_dd', ceplen=13, CMN='current', VARNORM='no', AGC='none' 75 | INFO: cmn.c(143): mean[0]= 12.00, mean[1..12]= 0.0 76 | INFO: acmod.c(171): Using subvector specification 0-12/13-25/26-38 77 | INFO: mdef.c(518): Reading model definition: cmusphinx-5prealpha-en-us-ptm-2.0//mdef 78 | INFO: bin_mdef.c(181): Allocating 142108 * 8 bytes (1110 KiB) for CD tree 79 | INFO: tmat.c(206): Reading HMM transition probability matrices: cmusphinx-5prealpha-en-us-ptm-2.0//transition_matrices 80 | INFO: acmod.c(124): Attempting to use PTM computation module 81 | INFO: ms_gauden.c(198): Reading mixture gaussian parameter: cmusphinx-5prealpha-en-us-ptm-2.0//means 82 | INFO: ms_gauden.c(292): 42 codebook, 3 feature, size: 83 | INFO: ms_gauden.c(294): 128x13 84 | INFO: ms_gauden.c(294): 128x13 85 | INFO: ms_gauden.c(294): 128x13 86 | INFO: ms_gauden.c(198): Reading mixture gaussian parameter: cmusphinx-5prealpha-en-us-ptm-2.0//variances 87 | INFO: ms_gauden.c(292): 42 codebook, 3 feature, size: 88 | INFO: ms_gauden.c(294): 128x13 89 | INFO: ms_gauden.c(294): 128x13 90 | INFO: ms_gauden.c(294): 128x13 91 | INFO: ms_gauden.c(354): 222 variance values floored 92 | INFO: ptm_mgau.c(476): Loading senones from dump file cmusphinx-5prealpha-en-us-ptm-2.0//sendump 93 | INFO: ptm_mgau.c(500): BEGIN FILE FORMAT DESCRIPTION 94 | INFO: ptm_mgau.c(563): Rows: 128, Columns: 5126 95 | INFO: ptm_mgau.c(595): Using memory-mapped I/O for senones 96 | INFO: ptm_mgau.c(835): Maximum top-N: 4 97 | INFO: phone_loop_search.c(115): State beam -225 Phone exit beam -225 Insertion penalty 0 98 | INFO: dict.c(320): Allocating 4285 * 20 bytes (83 KiB) for word entries 99 | INFO: dict.c(333): Reading main dictionary: dictionary.dic 100 | ERROR: "dict.c", line 205: Line 58: Failed to add the word 'GO' (duplicate?); ignored 101 | ERROR: "dict.c", line 205: Line 76: Failed to add the word 'KITCHEN' (duplicate?); ignored 102 | ERROR: "dict.c", line 205: Line 98: Failed to add the word 'ON' (duplicate?); ignored 103 | ERROR: "dict.c", line 205: Line 99: Failed to add the word 'ON(2)' (duplicate?); ignored 104 | ERROR: "dict.c", line 205: Line 145: Failed to add the word 'THE' (duplicate?); ignored 105 | ERROR: "dict.c", line 205: Line 146: Failed to add the word 'THE(2)' (duplicate?); ignored 106 | ERROR: "dict.c", line 205: Line 155: Failed to add the word 'TURN' (duplicate?); ignored 107 | INFO: dict.c(213): Allocated 0 KiB for strings, 1 KiB for phones 108 | INFO: dict.c(336): 177 words read 109 | INFO: dict.c(342): Reading filler dictionary: cmusphinx-5prealpha-en-us-ptm-2.0//noisedict 110 | INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones 111 | INFO: dict.c(345): 5 words read 112 | INFO: dict2pid.c(396): Building PID tables for dictionary 113 | INFO: dict2pid.c(406): Allocating 42^3 * 2 bytes (144 KiB) for word-initial triphones 114 | INFO: dict2pid.c(132): Allocated 21336 bytes (20 KiB) for word-final triphones 115 | INFO: dict2pid.c(196): Allocated 21336 bytes (20 KiB) for single-phone word triphones 116 | INFO: ngram_model_arpa.c(477): ngrams 1=136, 2=278, 3=295 117 | INFO: ngram_model_arpa.c(135): Reading unigrams 118 | INFO: ngram_model_arpa.c(516): 136 = #unigrams created 119 | INFO: ngram_model_arpa.c(195): Reading bigrams 120 | INFO: ngram_model_arpa.c(534): 278 = #bigrams created 121 | INFO: ngram_model_arpa.c(535): 36 = #prob2 entries 122 | INFO: ngram_model_arpa.c(543): 28 = #bo_wt2 entries 123 | INFO: ngram_model_arpa.c(292): Reading trigrams 124 | INFO: ngram_model_arpa.c(556): 295 = #trigrams created 125 | INFO: ngram_model_arpa.c(557): 21 = #prob3 entries 126 | INFO: ngram_search_fwdtree.c(99): 112 unique initial diphones 127 | INFO: ngram_search_fwdtree.c(148): 0 root, 0 non-root channels, 9 single-phone words 128 | INFO: ngram_search_fwdtree.c(186): Creating search tree 129 | INFO: ngram_search_fwdtree.c(192): before: 0 root, 0 non-root channels, 9 single-phone words 130 | INFO: ngram_search_fwdtree.c(326): after: max nonroot chan increased to 438 131 | INFO: ngram_search_fwdtree.c(339): after: 106 root, 310 non-root channels, 8 single-phone words 132 | INFO: ngram_search_fwdflat.c(157): fwdflat: min_ef_width = 4, max_sf_win = 25 133 | INFO: cmn_prior.c(131): cmn_prior_update: from < 40.00 3.00 -1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 > 134 | INFO: cmn_prior.c(149): cmn_prior_update: to < 47.56 7.43 -7.99 6.42 -2.02 3.66 5.76 4.74 0.96 -1.26 4.18 4.25 -0.48 > 135 | INFO: ngram_search_fwdtree.c(1553): 1759 words recognized (17/fr) 136 | INFO: ngram_search_fwdtree.c(1555): 111370 senones evaluated (1092/fr) 137 | INFO: ngram_search_fwdtree.c(1559): 75257 channels searched (737/fr), 10388 1st, 45022 last 138 | INFO: ngram_search_fwdtree.c(1562): 2693 words for which last channels evaluated (26/fr) 139 | INFO: ngram_search_fwdtree.c(1564): 5245 candidate words for entering last phone (51/fr) 140 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.04 CPU 2.000 xRT 141 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.53 wall 3.459 xRT 142 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 49 words 143 | INFO: ngram_search_fwdflat.c(945): 1458 words recognized (14/fr) 144 | INFO: ngram_search_fwdflat.c(947): 77514 senones evaluated (760/fr) 145 | INFO: ngram_search_fwdflat.c(949): 89975 channels searched (882/fr) 146 | INFO: ngram_search_fwdflat.c(951): 4294 words searched (42/fr) 147 | INFO: ngram_search_fwdflat.c(954): 2611 word transitions (25/fr) 148 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.27 CPU 1.245 xRT 149 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.27 wall 1.249 xRT 150 | INFO: ngram_search.c(1252): lattice start node .0 end node .61 151 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 152 | INFO: ngram_search.c(1383): Lattice has 233 nodes, 1120 links 153 | INFO: ps_lattice.c(1380): Bestpath score: -3452 154 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:61:100) = -204526 155 | INFO: ps_lattice.c(1441): Joint P(O,S) = -256660 P(S|O) = -52134 156 | INFO: ngram_search.c(874): bestpath 0.04 CPU 0.040 xRT 157 | INFO: ngram_search.c(877): bestpath 0.04 wall 0.039 xRT 158 | INFO: cmn_prior.c(131): cmn_prior_update: from < 47.56 7.43 -7.99 6.42 -2.02 3.66 5.76 4.74 0.96 -1.26 4.18 4.25 -0.48 > 159 | INFO: cmn_prior.c(149): cmn_prior_update: to < 38.10 1.46 -6.19 6.02 1.93 4.53 3.32 2.68 3.30 -0.91 1.89 1.68 1.42 > 160 | INFO: ngram_search_fwdtree.c(1553): 4037 words recognized (9/fr) 161 | INFO: ngram_search_fwdtree.c(1555): 383310 senones evaluated (846/fr) 162 | INFO: ngram_search_fwdtree.c(1559): 209592 channels searched (462/fr), 47581 1st, 91646 last 163 | INFO: ngram_search_fwdtree.c(1562): 6860 words for which last channels evaluated (15/fr) 164 | INFO: ngram_search_fwdtree.c(1564): 13732 candidate words for entering last phone (30/fr) 165 | INFO: ngram_search_fwdtree.c(1567): fwdtree 7.73 CPU 1.706 xRT 166 | INFO: ngram_search_fwdtree.c(1570): fwdtree 9.80 wall 2.162 xRT 167 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 59 words 168 | INFO: ngram_search_fwdflat.c(945): 2707 words recognized (6/fr) 169 | INFO: ngram_search_fwdflat.c(947): 183758 senones evaluated (406/fr) 170 | INFO: ngram_search_fwdflat.c(949): 166519 channels searched (367/fr) 171 | INFO: ngram_search_fwdflat.c(951): 8645 words searched (19/fr) 172 | INFO: ngram_search_fwdflat.c(954): 5352 word transitions (11/fr) 173 | INFO: ngram_search_fwdflat.c(957): fwdflat 3.49 CPU 0.770 xRT 174 | INFO: ngram_search_fwdflat.c(960): fwdflat 3.50 wall 0.773 xRT 175 | INFO: ngram_search.c(1252): lattice start node .0 end node .396 176 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 177 | INFO: ngram_search.c(1383): Lattice has 684 nodes, 2290 links 178 | INFO: ps_lattice.c(1380): Bestpath score: -9557 179 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:396:451) = -542438 180 | INFO: ps_lattice.c(1441): Joint P(O,S) = -622451 P(S|O) = -80013 181 | INFO: ngram_search.c(874): bestpath 0.12 CPU 0.027 xRT 182 | INFO: ngram_search.c(877): bestpath 0.12 wall 0.027 xRT 183 | INFO: cmn_prior.c(131): cmn_prior_update: from < 38.10 1.46 -6.19 6.02 1.93 4.53 3.32 2.68 3.30 -0.91 1.89 1.68 1.42 > 184 | INFO: cmn_prior.c(149): cmn_prior_update: to < 36.83 0.91 -6.31 5.62 2.25 4.15 3.74 3.16 3.27 -1.15 1.98 1.93 1.47 > 185 | INFO: ngram_search_fwdtree.c(1553): 427 words recognized (7/fr) 186 | INFO: ngram_search_fwdtree.c(1555): 47068 senones evaluated (724/fr) 187 | INFO: ngram_search_fwdtree.c(1559): 23840 channels searched (366/fr), 6466 1st, 8880 last 188 | INFO: ngram_search_fwdtree.c(1562): 819 words for which last channels evaluated (12/fr) 189 | INFO: ngram_search_fwdtree.c(1564): 1278 candidate words for entering last phone (19/fr) 190 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.05 CPU 1.615 xRT 191 | INFO: ngram_search_fwdtree.c(1570): fwdtree 1.50 wall 2.300 xRT 192 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 7 words 193 | INFO: ngram_search_fwdflat.c(945): 310 words recognized (5/fr) 194 | INFO: ngram_search_fwdflat.c(947): 8874 senones evaluated (137/fr) 195 | INFO: ngram_search_fwdflat.c(949): 5710 channels searched (87/fr) 196 | INFO: ngram_search_fwdflat.c(951): 560 words searched (8/fr) 197 | INFO: ngram_search_fwdflat.c(954): 245 word transitions (3/fr) 198 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.40 CPU 0.615 xRT 199 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.39 wall 0.607 xRT 200 | INFO: ngram_search.c(1252): lattice start node .0 end node .18 201 | INFO: ngram_search.c(1278): Eliminated 0 nodes before end node 202 | INFO: ngram_search.c(1383): Lattice has 85 nodes, 50 links 203 | INFO: ps_lattice.c(1380): Bestpath score: -737 204 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:18:63) = -54199 205 | INFO: ps_lattice.c(1441): Joint P(O,S) = -64265 P(S|O) = -10066 206 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 207 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.002 xRT 208 | INFO: cmn_prior.c(131): cmn_prior_update: from < 36.83 0.91 -6.31 5.62 2.25 4.15 3.74 3.16 3.27 -1.15 1.98 1.93 1.47 > 209 | INFO: cmn_prior.c(149): cmn_prior_update: to < 37.38 0.58 -6.03 5.45 2.22 4.48 4.18 2.98 3.42 -1.04 1.41 2.30 1.27 > 210 | INFO: ngram_search_fwdtree.c(1553): 2431 words recognized (15/fr) 211 | INFO: ngram_search_fwdtree.c(1555): 167258 senones evaluated (1026/fr) 212 | INFO: ngram_search_fwdtree.c(1559): 106731 channels searched (654/fr), 16854 1st, 60541 last 213 | INFO: ngram_search_fwdtree.c(1562): 3869 words for which last channels evaluated (23/fr) 214 | INFO: ngram_search_fwdtree.c(1564): 6044 candidate words for entering last phone (37/fr) 215 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.01 CPU 1.847 xRT 216 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.90 wall 2.395 xRT 217 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 57 words 218 | INFO: ngram_search_fwdflat.c(945): 1993 words recognized (12/fr) 219 | INFO: ngram_search_fwdflat.c(947): 117947 senones evaluated (724/fr) 220 | INFO: ngram_search_fwdflat.c(949): 127706 channels searched (783/fr) 221 | INFO: ngram_search_fwdflat.c(951): 6032 words searched (37/fr) 222 | INFO: ngram_search_fwdflat.c(954): 3775 word transitions (23/fr) 223 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.86 CPU 1.141 xRT 224 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.87 wall 1.146 xRT 225 | INFO: ngram_search.c(1252): lattice start node .0 end node .122 226 | INFO: ngram_search.c(1278): Eliminated 0 nodes before end node 227 | INFO: ngram_search.c(1383): Lattice has 336 nodes, 2112 links 228 | INFO: ps_lattice.c(1380): Bestpath score: -3385 229 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:122:161) = -210542 230 | INFO: ps_lattice.c(1441): Joint P(O,S) = -232416 P(S|O) = -21874 231 | INFO: ngram_search.c(874): bestpath 0.10 CPU 0.062 xRT 232 | INFO: ngram_search.c(877): bestpath 0.10 wall 0.065 xRT 233 | INFO: cmn_prior.c(99): cmn_prior_update: from < 37.38 0.58 -6.03 5.45 2.22 4.48 4.18 2.98 3.42 -1.04 1.41 2.30 1.27 > 234 | INFO: cmn_prior.c(116): cmn_prior_update: to < 37.47 0.49 -6.12 5.43 2.17 4.42 4.02 2.80 3.29 -1.08 1.54 2.34 1.22 > 235 | INFO: cmn_prior.c(131): cmn_prior_update: from < 37.47 0.49 -6.12 5.43 2.17 4.42 4.02 2.80 3.29 -1.08 1.54 2.34 1.22 > 236 | INFO: cmn_prior.c(149): cmn_prior_update: to < 40.21 1.15 -7.65 4.55 1.42 3.23 1.84 1.12 3.17 -0.79 2.61 3.34 1.49 > 237 | INFO: ngram_search_fwdtree.c(1553): 2475 words recognized (17/fr) 238 | INFO: ngram_search_fwdtree.c(1555): 160265 senones evaluated (1076/fr) 239 | INFO: ngram_search_fwdtree.c(1559): 102433 channels searched (687/fr), 15335 1st, 59595 last 240 | INFO: ngram_search_fwdtree.c(1562): 3699 words for which last channels evaluated (24/fr) 241 | INFO: ngram_search_fwdtree.c(1564): 6468 candidate words for entering last phone (43/fr) 242 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.79 CPU 1.872 xRT 243 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.68 wall 2.470 xRT 244 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 46 words 245 | INFO: ngram_search_fwdflat.c(945): 1866 words recognized (13/fr) 246 | INFO: ngram_search_fwdflat.c(947): 102750 senones evaluated (690/fr) 247 | INFO: ngram_search_fwdflat.c(949): 101429 channels searched (680/fr) 248 | INFO: ngram_search_fwdflat.c(951): 5113 words searched (34/fr) 249 | INFO: ngram_search_fwdflat.c(954): 3332 word transitions (22/fr) 250 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.65 CPU 1.107 xRT 251 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.66 wall 1.113 xRT 252 | INFO: ngram_search.c(1252): lattice start node .0 end node .110 253 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 254 | INFO: ngram_search.c(1383): Lattice has 347 nodes, 1487 links 255 | INFO: ps_lattice.c(1380): Bestpath score: -6142 256 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:110:147) = -375397 257 | INFO: ps_lattice.c(1441): Joint P(O,S) = -472330 P(S|O) = -96933 258 | INFO: ngram_search.c(874): bestpath 0.05 CPU 0.034 xRT 259 | INFO: ngram_search.c(877): bestpath 0.05 wall 0.031 xRT 260 | INFO: cmn_prior.c(99): cmn_prior_update: from < 40.21 1.15 -7.65 4.55 1.42 3.23 1.84 1.12 3.17 -0.79 2.61 3.34 1.49 > 261 | INFO: cmn_prior.c(116): cmn_prior_update: to < 40.16 1.33 -6.56 4.00 1.15 3.76 2.48 0.88 2.23 -1.34 2.00 2.99 1.68 > 262 | INFO: cmn_prior.c(131): cmn_prior_update: from < 40.16 1.33 -6.56 4.00 1.15 3.76 2.48 0.88 2.23 -1.34 2.00 2.99 1.68 > 263 | INFO: cmn_prior.c(149): cmn_prior_update: to < 39.79 1.37 -6.51 3.96 1.22 3.72 2.62 0.97 2.14 -1.35 2.06 2.99 1.75 > 264 | INFO: ngram_search_fwdtree.c(1553): 1672 words recognized (10/fr) 265 | INFO: ngram_search_fwdtree.c(1555): 144969 senones evaluated (853/fr) 266 | INFO: ngram_search_fwdtree.c(1559): 79508 channels searched (467/fr), 17522 1st, 36240 last 267 | INFO: ngram_search_fwdtree.c(1562): 2722 words for which last channels evaluated (16/fr) 268 | INFO: ngram_search_fwdtree.c(1564): 4502 candidate words for entering last phone (26/fr) 269 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.02 CPU 1.776 xRT 270 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.19 wall 3.054 xRT 271 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 37 words 272 | INFO: ngram_search_fwdflat.c(945): 1194 words recognized (7/fr) 273 | INFO: ngram_search_fwdflat.c(947): 68320 senones evaluated (402/fr) 274 | INFO: ngram_search_fwdflat.c(949): 63241 channels searched (372/fr) 275 | INFO: ngram_search_fwdflat.c(951): 3766 words searched (22/fr) 276 | INFO: ngram_search_fwdflat.c(954): 1803 word transitions (10/fr) 277 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.24 CPU 0.729 xRT 278 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.24 wall 0.730 xRT 279 | INFO: ngram_search.c(1252): lattice start node .0 end node .41 280 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 281 | INFO: ngram_search.c(1383): Lattice has 229 nodes, 117 links 282 | INFO: ps_lattice.c(1380): Bestpath score: -2867 283 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:41:168) = -249982 284 | INFO: ps_lattice.c(1441): Joint P(O,S) = -282789 P(S|O) = -32807 285 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.006 xRT 286 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.003 xRT 287 | INFO: cmn_prior.c(131): cmn_prior_update: from < 39.79 1.37 -6.51 3.96 1.22 3.72 2.62 0.97 2.14 -1.35 2.06 2.99 1.75 > 288 | INFO: cmn_prior.c(149): cmn_prior_update: to < 41.30 2.52 -5.95 3.73 0.82 3.56 3.04 0.43 2.12 -1.37 1.41 2.48 2.35 > 289 | INFO: ngram_search_fwdtree.c(1553): 1175 words recognized (12/fr) 290 | INFO: ngram_search_fwdtree.c(1555): 93778 senones evaluated (957/fr) 291 | INFO: ngram_search_fwdtree.c(1559): 58136 channels searched (593/fr), 9808 1st, 32006 last 292 | INFO: ngram_search_fwdtree.c(1562): 1950 words for which last channels evaluated (19/fr) 293 | INFO: ngram_search_fwdtree.c(1564): 3302 candidate words for entering last phone (33/fr) 294 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.87 CPU 1.908 xRT 295 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.59 wall 3.663 xRT 296 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 38 words 297 | INFO: ngram_search_fwdflat.c(945): 833 words recognized (8/fr) 298 | INFO: ngram_search_fwdflat.c(947): 61337 senones evaluated (626/fr) 299 | INFO: ngram_search_fwdflat.c(949): 61203 channels searched (624/fr) 300 | INFO: ngram_search_fwdflat.c(951): 2910 words searched (29/fr) 301 | INFO: ngram_search_fwdflat.c(954): 2161 word transitions (22/fr) 302 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.05 CPU 1.071 xRT 303 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.06 wall 1.079 xRT 304 | INFO: ngram_search.c(1252): lattice start node .0 end node .57 305 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 306 | INFO: ngram_search.c(1383): Lattice has 180 nodes, 142 links 307 | INFO: ps_lattice.c(1380): Bestpath score: -3123 308 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:57:96) = -232113 309 | INFO: ps_lattice.c(1441): Joint P(O,S) = -249738 P(S|O) = -17625 310 | INFO: ngram_search.c(874): bestpath 0.00 CPU 0.000 xRT 311 | INFO: ngram_search.c(877): bestpath 0.00 wall 0.004 xRT 312 | INFO: cmn_prior.c(131): cmn_prior_update: from < 41.30 2.52 -5.95 3.73 0.82 3.56 3.04 0.43 2.12 -1.37 1.41 2.48 2.35 > 313 | INFO: cmn_prior.c(149): cmn_prior_update: to < 42.07 1.79 -5.05 4.15 -1.11 5.68 2.16 1.48 2.68 -2.95 2.36 1.14 2.80 > 314 | INFO: ngram_search_fwdtree.c(1553): 2677 words recognized (16/fr) 315 | INFO: ngram_search_fwdtree.c(1555): 188245 senones evaluated (1114/fr) 316 | INFO: ngram_search_fwdtree.c(1559): 130753 channels searched (773/fr), 17490 1st, 81399 last 317 | INFO: ngram_search_fwdtree.c(1562): 4568 words for which last channels evaluated (27/fr) 318 | INFO: ngram_search_fwdtree.c(1564): 7793 candidate words for entering last phone (46/fr) 319 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.22 CPU 1.905 xRT 320 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.33 wall 2.564 xRT 321 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 67 words 322 | INFO: ngram_search_fwdflat.c(945): 1885 words recognized (11/fr) 323 | INFO: ngram_search_fwdflat.c(947): 131964 senones evaluated (781/fr) 324 | INFO: ngram_search_fwdflat.c(949): 151437 channels searched (896/fr) 325 | INFO: ngram_search_fwdflat.c(951): 6924 words searched (40/fr) 326 | INFO: ngram_search_fwdflat.c(954): 4727 word transitions (27/fr) 327 | INFO: ngram_search_fwdflat.c(957): fwdflat 2.04 CPU 1.207 xRT 328 | INFO: ngram_search_fwdflat.c(960): fwdflat 2.04 wall 1.205 xRT 329 | INFO: ngram_search.c(1252): lattice start node .0 end node .136 330 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 331 | INFO: ngram_search.c(1383): Lattice has 397 nodes, 2063 links 332 | INFO: ps_lattice.c(1380): Bestpath score: -6265 333 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:136:167) = -332527 334 | INFO: ps_lattice.c(1441): Joint P(O,S) = -388192 P(S|O) = -55665 335 | INFO: ngram_search.c(874): bestpath 0.08 CPU 0.048 xRT 336 | INFO: ngram_search.c(877): bestpath 0.08 wall 0.049 xRT 337 | INFO: cmn_prior.c(99): cmn_prior_update: from < 42.07 1.79 -5.05 4.15 -1.11 5.68 2.16 1.48 2.68 -2.95 2.36 1.14 2.80 > 338 | INFO: cmn_prior.c(116): cmn_prior_update: to < 42.24 1.69 -5.14 4.14 -1.46 5.72 2.45 1.21 2.59 -2.97 2.30 1.54 2.73 > 339 | INFO: cmn_prior.c(131): cmn_prior_update: from < 42.24 1.69 -5.14 4.14 -1.46 5.72 2.45 1.21 2.59 -2.97 2.30 1.54 2.73 > 340 | INFO: cmn_prior.c(149): cmn_prior_update: to < 44.12 1.97 -4.88 3.42 -2.21 5.14 2.00 0.32 2.84 -2.55 2.03 2.33 2.62 > 341 | INFO: ngram_search_fwdtree.c(1553): 2104 words recognized (19/fr) 342 | INFO: ngram_search_fwdtree.c(1555): 118003 senones evaluated (1044/fr) 343 | INFO: ngram_search_fwdtree.c(1559): 83428 channels searched (738/fr), 11530 1st, 53523 last 344 | INFO: ngram_search_fwdtree.c(1562): 3160 words for which last channels evaluated (27/fr) 345 | INFO: ngram_search_fwdtree.c(1564): 3340 candidate words for entering last phone (29/fr) 346 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.14 CPU 1.894 xRT 347 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.58 wall 3.168 xRT 348 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 74 words 349 | INFO: ngram_search_fwdflat.c(945): 1572 words recognized (14/fr) 350 | INFO: ngram_search_fwdflat.c(947): 111866 senones evaluated (990/fr) 351 | INFO: ngram_search_fwdflat.c(949): 129725 channels searched (1148/fr) 352 | INFO: ngram_search_fwdflat.c(951): 6313 words searched (55/fr) 353 | INFO: ngram_search_fwdflat.c(954): 4119 word transitions (36/fr) 354 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.54 CPU 1.363 xRT 355 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.54 wall 1.364 xRT 356 | INFO: ngram_search.c(1252): lattice start node .0 end node .69 357 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 358 | INFO: ngram_search.c(1383): Lattice has 245 nodes, 972 links 359 | INFO: ps_lattice.c(1380): Bestpath score: -5087 360 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:69:111) = -290318 361 | INFO: ps_lattice.c(1441): Joint P(O,S) = -321079 P(S|O) = -30761 362 | INFO: ngram_search.c(874): bestpath 0.03 CPU 0.027 xRT 363 | INFO: ngram_search.c(877): bestpath 0.03 wall 0.030 xRT 364 | INFO: cmn_prior.c(131): cmn_prior_update: from < 44.12 1.97 -4.88 3.42 -2.21 5.14 2.00 0.32 2.84 -2.55 2.03 2.33 2.62 > 365 | INFO: cmn_prior.c(149): cmn_prior_update: to < 45.89 2.47 -4.85 4.90 -2.78 5.25 2.23 -0.23 2.56 -2.15 2.43 2.72 2.53 > 366 | INFO: ngram_search_fwdtree.c(1553): 1616 words recognized (15/fr) 367 | INFO: ngram_search_fwdtree.c(1555): 99964 senones evaluated (926/fr) 368 | INFO: ngram_search_fwdtree.c(1559): 64624 channels searched (598/fr), 11006 1st, 38670 last 369 | INFO: ngram_search_fwdtree.c(1562): 2401 words for which last channels evaluated (22/fr) 370 | INFO: ngram_search_fwdtree.c(1564): 2483 candidate words for entering last phone (22/fr) 371 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.07 CPU 1.917 xRT 372 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.25 wall 3.939 xRT 373 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 49 words 374 | INFO: ngram_search_fwdflat.c(945): 1335 words recognized (12/fr) 375 | INFO: ngram_search_fwdflat.c(947): 82171 senones evaluated (761/fr) 376 | INFO: ngram_search_fwdflat.c(949): 84855 channels searched (785/fr) 377 | INFO: ngram_search_fwdflat.c(951): 4356 words searched (40/fr) 378 | INFO: ngram_search_fwdflat.c(954): 2710 word transitions (25/fr) 379 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.25 CPU 1.157 xRT 380 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.25 wall 1.158 xRT 381 | INFO: ngram_search.c(1252): lattice start node .0 end node .75 382 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 383 | INFO: ngram_search.c(1383): Lattice has 228 nodes, 997 links 384 | INFO: ps_lattice.c(1380): Bestpath score: -4891 385 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:75:106) = -303625 386 | INFO: ps_lattice.c(1441): Joint P(O,S) = -336280 P(S|O) = -32655 387 | INFO: ngram_search.c(874): bestpath 0.03 CPU 0.028 xRT 388 | INFO: ngram_search.c(877): bestpath 0.03 wall 0.029 xRT 389 | INFO: cmn_prior.c(99): cmn_prior_update: from < 45.89 2.47 -4.85 4.90 -2.78 5.25 2.23 -0.23 2.56 -2.15 2.43 2.72 2.53 > 390 | INFO: cmn_prior.c(116): cmn_prior_update: to < 46.75 2.19 -4.66 7.17 -1.79 5.31 2.88 -0.01 2.93 -2.55 1.81 2.89 2.84 > 391 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.75 2.19 -4.66 7.17 -1.79 5.31 2.88 -0.01 2.93 -2.55 1.81 2.89 2.84 > 392 | INFO: cmn_prior.c(149): cmn_prior_update: to < 46.62 2.18 -4.68 7.21 -1.75 5.33 2.84 0.02 2.97 -2.48 1.76 2.92 2.87 > 393 | INFO: ngram_search_fwdtree.c(1553): 1212 words recognized (12/fr) 394 | INFO: ngram_search_fwdtree.c(1555): 93356 senones evaluated (943/fr) 395 | INFO: ngram_search_fwdtree.c(1559): 57607 channels searched (581/fr), 10069 1st, 32592 last 396 | INFO: ngram_search_fwdtree.c(1562): 1963 words for which last channels evaluated (19/fr) 397 | INFO: ngram_search_fwdtree.c(1564): 2985 candidate words for entering last phone (30/fr) 398 | INFO: ngram_search_fwdtree.c(1567): fwdtree 1.86 CPU 1.879 xRT 399 | INFO: ngram_search_fwdtree.c(1570): fwdtree 3.65 wall 3.683 xRT 400 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 29 words 401 | INFO: ngram_search_fwdflat.c(945): 1046 words recognized (11/fr) 402 | INFO: ngram_search_fwdflat.c(947): 57250 senones evaluated (578/fr) 403 | INFO: ngram_search_fwdflat.c(949): 49789 channels searched (502/fr) 404 | INFO: ngram_search_fwdflat.c(951): 2641 words searched (26/fr) 405 | INFO: ngram_search_fwdflat.c(954): 1652 word transitions (16/fr) 406 | INFO: ngram_search_fwdflat.c(957): fwdflat 0.90 CPU 0.909 xRT 407 | INFO: ngram_search_fwdflat.c(960): fwdflat 0.91 wall 0.915 xRT 408 | INFO: ngram_search.c(1252): lattice start node .0 end node .57 409 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 410 | INFO: ngram_search.c(1383): Lattice has 157 nodes, 191 links 411 | INFO: ps_lattice.c(1380): Bestpath score: -2234 412 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:57:97) = -185629 413 | INFO: ps_lattice.c(1441): Joint P(O,S) = -191948 P(S|O) = -6319 414 | INFO: ngram_search.c(874): bestpath 0.01 CPU 0.010 xRT 415 | INFO: ngram_search.c(877): bestpath 0.01 wall 0.007 xRT 416 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.62 2.18 -4.68 7.21 -1.75 5.33 2.84 0.02 2.97 -2.48 1.76 2.92 2.87 > 417 | INFO: cmn_prior.c(149): cmn_prior_update: to < 47.20 3.10 -6.90 6.47 -1.18 5.67 2.73 -0.79 2.81 -1.68 1.80 3.73 3.10 > 418 | INFO: ngram_search_fwdtree.c(1553): 2107 words recognized (19/fr) 419 | INFO: ngram_search_fwdtree.c(1555): 122471 senones evaluated (1084/fr) 420 | INFO: ngram_search_fwdtree.c(1559): 86113 channels searched (762/fr), 11551 1st, 53541 last 421 | INFO: ngram_search_fwdtree.c(1562): 3073 words for which last channels evaluated (27/fr) 422 | INFO: ngram_search_fwdtree.c(1564): 4614 candidate words for entering last phone (40/fr) 423 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.27 CPU 2.009 xRT 424 | INFO: ngram_search_fwdtree.c(1570): fwdtree 4.47 wall 3.952 xRT 425 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 70 words 426 | INFO: ngram_search_fwdflat.c(945): 1973 words recognized (17/fr) 427 | INFO: ngram_search_fwdflat.c(947): 117658 senones evaluated (1041/fr) 428 | INFO: ngram_search_fwdflat.c(949): 134783 channels searched (1192/fr) 429 | INFO: ngram_search_fwdflat.c(951): 6480 words searched (57/fr) 430 | INFO: ngram_search_fwdflat.c(954): 4226 word transitions (37/fr) 431 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.68 CPU 1.487 xRT 432 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.68 wall 1.490 xRT 433 | INFO: ngram_search.c(1252): lattice start node .0 end node .80 434 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 435 | INFO: ngram_search.c(1383): Lattice has 274 nodes, 1536 links 436 | INFO: ps_lattice.c(1380): Bestpath score: -4830 437 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:80:111) = -266760 438 | INFO: ps_lattice.c(1441): Joint P(O,S) = -316158 P(S|O) = -49398 439 | INFO: ngram_search.c(874): bestpath 0.06 CPU 0.054 xRT 440 | INFO: ngram_search.c(877): bestpath 0.06 wall 0.054 xRT 441 | INFO: cmn_prior.c(99): cmn_prior_update: from < 47.20 3.10 -6.90 6.47 -1.18 5.67 2.73 -0.79 2.81 -1.68 1.80 3.73 3.10 > 442 | INFO: cmn_prior.c(116): cmn_prior_update: to < 47.76 2.30 -7.04 5.35 -2.54 6.02 2.14 -0.56 3.12 -1.75 3.01 2.41 2.77 > 443 | INFO: cmn_prior.c(131): cmn_prior_update: from < 47.76 2.30 -7.04 5.35 -2.54 6.02 2.14 -0.56 3.12 -1.75 3.01 2.41 2.77 > 444 | INFO: cmn_prior.c(149): cmn_prior_update: to < 46.82 1.84 -6.64 5.20 -2.52 6.06 2.61 -0.52 3.11 -1.63 2.76 2.36 2.80 > 445 | INFO: ngram_search_fwdtree.c(1553): 2902 words recognized (14/fr) 446 | INFO: ngram_search_fwdtree.c(1555): 203842 senones evaluated (1004/fr) 447 | INFO: ngram_search_fwdtree.c(1559): 135011 channels searched (665/fr), 20153 1st, 81601 last 448 | INFO: ngram_search_fwdtree.c(1562): 4870 words for which last channels evaluated (23/fr) 449 | INFO: ngram_search_fwdtree.c(1564): 7434 candidate words for entering last phone (36/fr) 450 | INFO: ngram_search_fwdtree.c(1567): fwdtree 3.80 CPU 1.872 xRT 451 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.16 wall 2.544 xRT 452 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 66 words 453 | INFO: ngram_search_fwdflat.c(945): 2195 words recognized (11/fr) 454 | INFO: ngram_search_fwdflat.c(947): 147318 senones evaluated (726/fr) 455 | INFO: ngram_search_fwdflat.c(949): 158766 channels searched (782/fr) 456 | INFO: ngram_search_fwdflat.c(951): 7892 words searched (38/fr) 457 | INFO: ngram_search_fwdflat.c(954): 4849 word transitions (23/fr) 458 | INFO: ngram_search_fwdflat.c(957): fwdflat 2.33 CPU 1.148 xRT 459 | INFO: ngram_search_fwdflat.c(960): fwdflat 2.34 wall 1.151 xRT 460 | INFO: ngram_search.c(1252): lattice start node .0 end node .161 461 | INFO: ngram_search.c(1278): Eliminated 2 nodes before end node 462 | INFO: ngram_search.c(1383): Lattice has 364 nodes, 1874 links 463 | INFO: ps_lattice.c(1380): Bestpath score: -8460 464 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:161:201) = -448872 465 | INFO: ps_lattice.c(1441): Joint P(O,S) = -517908 P(S|O) = -69036 466 | INFO: ngram_search.c(874): bestpath 0.07 CPU 0.035 xRT 467 | INFO: ngram_search.c(877): bestpath 0.07 wall 0.033 xRT 468 | INFO: cmn_prior.c(131): cmn_prior_update: from < 46.82 1.84 -6.64 5.20 -2.52 6.06 2.61 -0.52 3.11 -1.63 2.76 2.36 2.80 > 469 | INFO: cmn_prior.c(149): cmn_prior_update: to < 48.32 2.52 -7.42 4.72 -2.46 5.29 2.11 -0.73 2.93 -0.99 2.85 3.02 2.62 > 470 | INFO: ngram_search_fwdtree.c(1553): 1771 words recognized (15/fr) 471 | INFO: ngram_search_fwdtree.c(1555): 108611 senones evaluated (890/fr) 472 | INFO: ngram_search_fwdtree.c(1559): 74406 channels searched (609/fr), 12308 1st, 46058 last 473 | INFO: ngram_search_fwdtree.c(1562): 2796 words for which last channels evaluated (22/fr) 474 | INFO: ngram_search_fwdtree.c(1564): 3011 candidate words for entering last phone (24/fr) 475 | INFO: ngram_search_fwdtree.c(1567): fwdtree 2.42 CPU 1.984 xRT 476 | INFO: ngram_search_fwdtree.c(1570): fwdtree 5.54 wall 4.544 xRT 477 | INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 60 words 478 | INFO: ngram_search_fwdflat.c(945): 1364 words recognized (11/fr) 479 | INFO: ngram_search_fwdflat.c(947): 94821 senones evaluated (777/fr) 480 | INFO: ngram_search_fwdflat.c(949): 111967 channels searched (917/fr) 481 | INFO: ngram_search_fwdflat.c(951): 5288 words searched (43/fr) 482 | INFO: ngram_search_fwdflat.c(954): 3632 word transitions (29/fr) 483 | INFO: ngram_search_fwdflat.c(957): fwdflat 1.58 CPU 1.295 xRT 484 | INFO: ngram_search_fwdflat.c(960): fwdflat 1.59 wall 1.306 xRT 485 | INFO: ngram_search.c(1252): lattice start node .0 end node .90 486 | INFO: ngram_search.c(1278): Eliminated 1 nodes before end node 487 | INFO: ngram_search.c(1383): Lattice has 244 nodes, 1181 links 488 | INFO: ps_lattice.c(1380): Bestpath score: -5640 489 | INFO: ps_lattice.c(1384): Normalizer P(O) = alpha(:90:120) = -320760 490 | INFO: ps_lattice.c(1441): Joint P(O,S) = -378437 P(S|O) = -57677 491 | INFO: ngram_search.c(874): bestpath 0.04 CPU 0.033 xRT 492 | INFO: ngram_search.c(877): bestpath 0.04 wall 0.034 xRT 493 | INFO: ngram_search_fwdtree.c(432): TOTAL fwdtree 39.29 CPU 1.859 xRT 494 | INFO: ngram_search_fwdtree.c(435): TOTAL fwdtree 62.17 wall 2.942 xRT 495 | INFO: ngram_search_fwdflat.c(176): TOTAL fwdflat 22.28 CPU 1.054 xRT 496 | INFO: ngram_search_fwdflat.c(179): TOTAL fwdflat 22.34 wall 1.057 xRT 497 | INFO: ngram_search.c(303): TOTAL bestpath 0.64 CPU 0.030 xRT 498 | INFO: ngram_search.c(306): TOTAL bestpath 0.64 wall 0.030 xRT 499 | --------------------------------------------------------------------------------