Speech:Spring 2016 Meagan Wolf Log


 * Home
 * Semesters
 * Spring 2016
 * Proposal
 * Report
 * Information - General Project Information
 * Experiments - List of speech experiments

Week Ending February 9, 2016
I was supposed to log in and explore the directories and if I felt comfortable enough run a train. .   I had a migraine on Thursday, Friday, and Saturday, and then Sunday I felt as if I had a fever and until Tuesday night had no energy.
 * Task:

While I couldn't log in I did read through a good deal of documentation and made some lists for myself of steps that I needed to make sure that I remember (like how to get onto Caesar). I did log in under root for a brief period of time to explore the directories. I didn't get a chance to run anything because I didn't feel comfortable enough yet.
 * Results:

1. Learn how to log in to Caesar
 * Plan:

2. Become familiar with the coding language being used in Caesar
 * Concerns:

Week Ending February 16, 2016
My groups task for this week is to create a sub directory under 0282 and number it 004. Before I started to run a train I wanted to make sure that I knew exactly how to run the train so that I could do my part, so I communicated with Matthew through email. I am currently logged on trying to figure out how to run a train. So far it is going successfully.
 * Task:

2/10: I communicated with Matthew through email so that I could make sure I had all the correct steps to successfully run a train without screwing anything up.
 * Results:

2/15: I communicated with Matthew more to make sure I knew which sub directory to create and where to create it.

2/16: I logged on and tried running my first test. So far it seems like it's going well, but there are a few results that I'm not sure if that's what it should be producing.

I'm going to try to get on tomorrow before class and try to run another train, just so that I can become more comfortable with the process of how to do it.
 * Plan:

I'm still not feeling 100% confident with the language, but I think if I can get a lot more practice running trains and just using it I will slowly but surely start to feel comfortable. The other challenging part is that I need to make sure I understand how to do whatever the task for the week is, so that I don't have to ask a group member for help the day after class and then have to wait for a response, this cuts into the amount of time I can spend actually logged in doing work.
 * Concerns:

Week Ending February 23, 2016
2/20: I know that we are supposed to work on individual tasks. As of right now I don't know what my part is, so I have emailed Matthew to find that out. Hopefully I'll hear from him later on today, so I can get some work done. I will be researching how the training and decoding process works, so that when we begin creating the prepareDecode.pl script I am knowledgeable about the process. 2/21: We just finished out group meeting and I now know that I need to create the software repository and organize the experiment directory. To organize the experiment I need to figure out which experiments are being used and which ones aren't. With the experiments that aren't being used they will be moved somewhere different or they will be deleted. 2/22: I've been researching to make sure I know exactly what I need to do and the commands that I need to use to accomplish my assigned task, so that tomorrow I can get into Caesar and make note of the files that are there and also which files are being used and which ones aren't. 2/23: Today I'm going to login to Caesar and explore what files are in the scripts/user directory, so that we (as a group can discuss this when we meet next) and then find out how I should best organize this directory and also if we need to delete any files, like if there are any that aren't being used at all.
 * Task:

2/20: I started reading through existing documentation on this site to get a better idea of how the training and decoding processes are done. 2/21: During the meeting I took notes on what I my assigned task was, but I forgot to write down the Perl commands necessary to see all the files in the scripts/user directory and the important information that goes with those files, so I emailed Matthew. 2/22: While I was waiting for Matthew to get back to me on the Perl commands that I would need to use I did some research to see if I could find out what the command was. 2/23: I successfully logged in to Caesar to explore what files are in the scripts/user directory, so that as a group we can discuss what the next step in organizing this directory. The following are all the files in the scripts/user directory: total 372 drwxr-xr-x. 11 root   root    4096 Feb 21 19:59 History -rwxr-xr-x. 1   2412 cis790  6288 Feb  4 17:00 createWiki_Sub_Experiment.pl -rwxr-xr-x. 1 root   root    2997 Feb  3 16:12 prepareTrainExperiment.pl -rwxrwxrwx. 1   2418 cis790  4473 Feb  3 16:07 genTrans.pl -rwxrwxr-x. 1 root   cis790  9832 Feb  3 15:49 setup_SphinxTrain.pl -rw-r--r--. 1 mcy59  cis790     0 Jan 27 16:31 new -rw-r--r--. 1 mcy59  cis790     0 Jan 27 16:31 clear -rwxr-xr-x. 1 root   root    2997 Jul 24  2015 prepareDecode.pl -rwxr-xr-x. 1 root   root    2997 Jul 21  2015 7212015prepareExperiment.pl -rwxrwxrwx. 1 root   root     375 Jul 15  2015 generateFeats.pl -rw-r--r--. 1   2303 cis790  4982 Jul  3  2015 exp_dir_setup.pl -rwxr-xr-x. 1 root   root    5467 Jul  3  2015 createWikiExperiment.pl -rwxr-xr-x. 1 root   root     762 Apr  6  2015 run_decode.pl drwxr-xr-x. 5 root   root    4096 Mar 27  2015 DELETE -rwxr-xr-x. 1 root   root    1405 Mar 11  2015 pruneDictionary.pl -rwxrwxrwx. 1   2418 cis790   303 Mar  4  2015 insertSil.pl -rwxrwxrwx. 1   2405 cis790 30270 Feb 25  2015 master_run_train.pl -rwxrwxrwx. 1   2303 cis790  4368 Jan 18  2015 exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790  1003 Jun 10  2014 ParseTranscript2.pl -rwxrwxr-x. 1 root   cis790  1716 Jun 10  2014 addQueueUsers.pl -rwxrwxrwx. 1   2303 cis790  6674 Jun 10  2014 child_exp_dir_setup.pl -rwxrwxrwx. 1   2303 cis790  5233 Jun 10  2014 child_exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790   900 Jun 10  2014 convert.pl -rwxrwxr-x. 1 root   cis790  1912 Jun 10  2014 copySph.pl -rwxrwxr-x. 1   2310 cis790   743 Jun 10  2014 corpusSize.pl -rwxrwxr-x. 1 root   cis790  2152 Jun 10  2014 createTranscript.pl -rwxrwxr-x. 1 root   cis790   729 Jun 10  2014 createdict.pl -rwxrwxr-x. 1 root   cis790   182 Jun 10  2014 find.pl -rwxrwxr-x. 1 root   cis790    88 Jun 10  2014 genFileIDs.csh -rwxrwxr-x. 1 root   cis790   118 Jun 10  2014 genPhones.csh -rwxrwxr-x. 1 root   cis790   956 Jun 10  2014 lm_create.pl -rwxrwxr-x. 1 root   cis790  1199 Jun 10  2014 parseDecode.pl -rw-r--r--. 1 root   root    2658 Apr 27  2014 monoGen2.pl -rwxr-xr-x. 1   2310 cis790  2339 Apr 22  2014 extract23.pl -rwxr-xr-x. 1 root   root     938 Apr 21  2014 copySph3.pl -rwxr-xr-x. 1 root   root    1405 Apr 11  2014 buildData2.pl -rwxr-xr-x. 1 root   root     940 Apr 11  2014 copySph2.pl -rw-r--r--. 1 root   root     159 Apr 10  2014 updateCFG.sh -rwxr-xr-x. 1 root   root    2860 Apr 10  2014 buildData.pl -rwxr-xr-x. 1 frs2000 cis790 9764 Apr  8  2014 fs_setup_SphinxTrain.pl -rwxr-xr-x. 1 root   root     577 Apr  6  2014 corpusSize0.pl -rwxr-xr-x. 1   2311 cis790   520 Mar 28  2014 trans_time.pl -rw-r--r--. 1   2311 cis790   959 Mar 26  2014 checkTrain.pl -rwxr-xr-x. 1   2310 cis790   915 Mar  4  2014 createSubTranscript.pl -rwxr-xr-x. 1   2310 cis790   838 Mar  4  2014 corpusSize2.pl -rw-r--r--. 1   2310 cis790   123 Mar  3  2014 cleanTrans.sh -rw-r--r--. 1 root   root   55153 Feb 28  2014 buildme.sh -rwxr--r-x. 1   2303 cis790  9970 Feb 26  2014 master_run_train_old.pl -rw-r--r--. 1   2310 cis790   450 Feb 26  2014 pullFromTrans.pl -rw-r--r--. 1   2310 cis790   719 Feb 13  2014 gen_errors.pl -rwxr-xr-x. 1   2310 cis790  2765 Feb  6  2014 dictionary3.pl -rwxr-xr-x. 1   2204 cis790 15377 Aug 31  2013 awesomeAdd.pl -rwxr-xr-x. 1   2204 cis790  2765 Aug 31  2013 dictionary2.pl -rwxr-xr-x. 1   2204 cis790  5300 Aug  6  2013 clone_exp.pl -rwxr-xr-x. 1   2204 cis790  6997 Aug  6  2013 train_02.pl -rwxr-xr-x. 1   2204 cis790  7731 Aug  6  2013 train_01.pl
 * Results:

2/20: The plan for this week is that we are supposed to work on individual tasks throughout the week. My task is to research how the training and decoding process works, so that when we (as a group) begin creating the prepareDecode.pl script I am knowledgeable about the process. In addition to this Sunday evening at 7PM we will be holding a Skype meeting to work on the proposal. 2/21: We just finished out group meeting and I now know that I need to create the software repository and organize the experiment directory. To organize the experiment I need to figure out which experiments are being used and which ones aren't. With the experiments that aren't being used they will be moved somewhere different or they will be deleted. 2/22: I'm making sure I know exactly what I need to do and the commands that I need to use to accomplish my assigned task, so that tomorrow I can get into Caesar and make note of the files that are there and also which files are being used and which ones aren't. 2/23: I successfully logged in to Caesar to explore what files are in the scripts/user directory, so that we (as a group can discuss this when we meet next) and then find out how I should best organize this directory and also if we need to delete any files, like if there are any that aren't being used at all. 2/20: I had no concerns then because I didn't find out my exact task until Sunday morning. 2/21: My concerns now that I know what the exact task that I will be doing is that my task of creating the software repository and organizing the experiment directory involve using Perl commands that I've never used before. I'm kind of scared that because I've never used these commands that I'll screw something up and that's the last thing that I want to do. 2/22: My concerns were how am I going to get everything done when I don't know where to start, so I decided to go and read documentation, to see if there was another group that did something similar to give me an idea of what commands I would need to use. 2/23: My concerns are now in regards to how I would go and explore in more depth each of the files in the scripts/users directory. I'd ideally want to read a description of what each of the scripts do and weather they are currently in use or not.
 * Plan:
 * Concerns:

Week Ending March 1, 2016
2/24: During class today I worked on going through all of the files in the scripts/users directory and comparing them with the what's listed on the Scripts Wiki page and making notes of what I found out. Some of my findings are that there are a good number of files in the directory that are not documented at all on the Scripts Wiki page. There are also multiple copies of the same file and it doesn't say what was updated. In general there are a lot of discrepancies between the Scripts Wiki page and the scripts/users directory. From the information that I find figure out what the best way to organize the scripts. 2/28: My tasks for this week are to answer some of the questions that I made note of by doing some research into the scripts and updating my notes. I should also create a document that's separate from my wiki page with an explanation of the important scripts. 2/29: I'm trying to find answers why in the scripts/users directory some of the file names show up in green, some show up in blue, and there are others that don't show up in color. 3/2: I still need to figure what each of the important scripts do and write a brief description, but that's a task to accomplish next this upcoming week.
 * Task:

2/24: In the directory there are some file names that show up in green, some that show up in blue, and others that don't show up in color. I don't know if that means anything. buildData.pl (April 10 2014) On the Scripts page it says that it was new for 2014 and next to that it says obsolete. buildData2.pl (April 11 2014) On the scripts page there's no record that this was created checkTrain.pl (Mar 26 2014) In the directory this isn't highlighted and colored. I don't know why, but I will find out clone_exp.pl (Aug 6 2013) convert.pl (Jun 10 2014) copySph.pl (Jun 10 2014) There are copySph2.pl and copySph3.pl. I don't yet know what the difference is between these two but will find that out. createTranscript.pl (Jun 10 2014) createSubTranscript.pl (Mar 4 2014) dictionary.pl (can't find in directory) Scripts log says this is the same as createdict.pl (Jun 10 2014) There's also dictionary2.pl and dictionary3.pl will explore more and see what they do and what the difference between them is. find.pl (Jun 10 2014) gen_errors.pl (Feb 13 2013) I don't know if it means something that this file isn't colored green or blue? generateFeats.pl (Jul 15 2015) genFileIDs.csh (Jun 10 2014) genTrans.pl (Feb 3 3016) The Scripts log says there should be multiple versions for this file, but I can only find this one. There's no record that in the log that they were deleted or something. lm_create.pl (Jun 10 2014) master_run_train.pl (Feb 25 2014) In the scripts/user directory there's record of a master_run_train_old.pl. In the Scripts log there's no record of this file, so I will try and find out what this does and if it's different from the master_run_train.pl parseDecode.pl (Jun 10 2014) 7212014prepareExperiment.pl (Seems like this is the most recent one). I don't know why it was named the way it is when it is the only prepareExperiment file in the scripts/user directory. So maybe eventually rename this file. pruneDictionary (Mar 11 2015) train_01.pl (Aug 6 2013) train_02.pl (Aug 6 2013) trans_time.pl (Mar 28 2014) updateDict (It's listed in the Scripts log in the Wiki, but I can't find in scripts/user directory) createWikiExperiment.pl (Jul 3 2015) createWiki_Sub_Experiment.pl (Feb 4 2015) There are some other Scripts that are in the scripts/user directory, but not documented or listed in the Scripts log on the Wiki page setup_SphinxTrain.pl (Feb 3 2016) exp_dir_setup.pl (Jul 3 2015) This is another file that's not colored green or blue. But I don't know what it does. run_decode.pl (Apr 6 2015) insertSil.pl (Mar 4 2015) exp_sphinx_config.pl (Jan 18 2015) addQueUsers.pl (Jun 10 2014) child_exp_dir_setup.pl (Jun 10 2014) child_exp_sphinx_config.pl (Jun 20 2014) corpusSize.pl (Jun 10 2014) genPhones.csh (Jun 10 2014) monoGen2.pl (Apr 27 2014) extract23.pl (Apr 22 2014) updateCFG.sh (Apr 11 2014) fs_setup_SphinxTrain.pl (Apr 10 2014) corpusSize0.pl (Apr 6 2014) corpusSize2.pl (Mar 4 2014) cleanTrans.sh (Mar 3 2014) buildme.sh (Feb 28 2014) pullFromTrans.pl (Feb 26 2014) awesomeAdd.pl (Aug 31 2013)
 * Results:

2/29: I'm trying to find answers why in the scripts/users directory some of the file names show up in green, some show up in blue, and there are others that don't show up in color. Just by going and logging into Caesar and changing directory to the scripts/users directory I think the file names in green are the ones that are currently in use or they could be the most recently created or edited scripts. I think the ones that show up in blue are other directories (like the DELETE) that are available from the scripts/users directory because they don't show up like normal scripts. I'm thinking that the scripts that aren't colored are ones that might not currently be used or they could be the scripts that were created a while ago.

2/24: During class today I worked on going through all of the files in the scripts/users directory and comparing them with the what's listed on the Scripts Wiki page and making notes of what I found out. Some of my findings are that there are a good number of files in the directory that are not documented at all on the Scripts Wiki page. There are also multiple copies of the same file and it doesn't say what was updated. In general there are a lot of discrepancies between the Scripts Wiki page and the scripts/users directory. From the information that I find figure out what the best way to organize the scripts. 2/28: We are doing either a Google Hangout or a Skype meeting at 6:30. My tasks for this week are to answer some of the questions that I made note of by doing some research into the scripts and updating my notes. I should also create a document that's separate from my wiki page with an explanation of the important scripts. 2/29: I'm trying to find answers why in the scripts/users directory some of the file names show up in green, some show up in blue, and there are others that don't show up in color. Just by going and logging into Caesar and changing directory to the scripts/users directory I think the file names in green are the ones that are currently in use or they could be the most recently created or edited scripts. I think the ones that show up in blue are other directories (like the DELETE) that are available from the scripts/users directory because they don't show up like normal scripts. I'm thinking that the scripts that aren't colored are ones that might not currently be used or they could be the scripts that were created a while ago. 3/2: I still need to figure what each of the important scripts do and write a brief description. 2/24: Right now my biggest concern is that we (as a group) aren't holding our outside of class meeting until Saturday afternoon. So I don't really have a lot to do between now and then. 2/28: Right now I don't really have any concerns. 2/29: I don't have any concerns as of right now.
 * Plan:
 * Concerns:

Week Ending March 8, 2016
3/2: Today during class I worked on writing a brief description of what the scripts in the directory do, so that we can go through and update the Scripts page on the Wiki. 3/7: Even though I haven't gotten a chance to work on writing descriptions of more of the scripts in the scripts/users directory I'm planning on doing more tomorrow and then Wednesday I will hopefully feel comfortable enough that I won't screw anything up by running any of the scripts that are in the directory but aren't documented. 3/8: I went through the rest of the scripts in the scripts/users and wrote descriptions of them. There were some that I found that haven't been documented on the Scripts Wiki page that I could easily get a description of what they are used for and the ones that I couldn't figure out what they do they gave me permission denied. 3/9: I just wanted to post my findings from going through the Wiki Scripts page and running some of the scripts that weren't documented that I could. I found that some of the issues with some of the scripts being documented on the scripts page and not being in the scripts/users directory was fixed. There were some other scripts that were in the scripts/users directory that weren't documented on the Wiki Scripts page.
 * Task:

3/2: Today during class I worked on writing a brief description of what the scripts in the directory do, so that we can go through and update the Scripts page on the Wiki. 3/7: Even though I haven't gotten a chance to work on writing descriptions of more of the scripts in the scripts/users directory I'm planning on doing more tomorrow and then Wednesday I will hopefully feel comfortable enough that I won't screw anything up by running any of the scripts that are in the directory but aren't documented. 3/8: I went through the rest of the scripts in the scripts/users and wrote descriptions of them. There were some that I found that haven't been documented on the Scripts Wiki page that I could easily get a description of what they are used for and the ones that I couldn't figure out what they do they gave me permission denied. while I was doing this I found out that there are less scripts in the scripts/users directory. total 312 drwxr-xr-x. 23 root   root    4096 Mar  5 10:15 History -rwxr-xr-x. 1 root   root     743 Mar  5 10:00 corpusSize.pl -rwxrwxrwx. 1 root   root     410 Mar  3 19:08 generateFeats.pl -rwxr-xr-x. 1   2412 cis790  6288 Feb  4 17:00 createWiki_Sub_Experiment.pl -rwxr-xr-x. 1 root   root    2997 Feb  3 16:12 prepareTrainExperiment.pl -rwxrwxrwx. 1   2418 cis790  4473 Feb  3 16:07 genTrans.pl -rwxrwxr-x. 1 root   cis790  9832 Feb  3 15:49 setup_SphinxTrain.pl -rw-r--r--. 1 mcy59  cis790     0 Jan 27 16:31 clear -rwxr-xr-x. 1 root   root    2997 Jul 24  2015 prepareDecode.pl -rwxr-xr-x. 1   2303 cis790  4982 Jul  3  2015 exp_dir_setup.pl -rwxr-xr-x. 1 root   root    5467 Jul  3  2015 createWikiExperiment.pl -rwxr-xr-x. 1 root   root     762 Apr  6  2015 run_decode.pl -rwxr-xr-x. 1 root   root    1405 Mar 11  2015 pruneDictionary.pl -rwxrwxrwx. 1   2418 cis790   303 Mar  4  2015 insertSil.pl -rwxrwxrwx. 1   2405 cis790 30270 Feb 25  2015 master_run_train.pl -rwxrwxrwx. 1   2303 cis790  4368 Jan 18  2015 exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790  1003 Jun 10  2014 ParseTranscript.pl -rwxrwxr-x. 1 root   cis790  1716 Jun 10  2014 addQueueUsers.pl -rwxrwxrwx. 1   2303 cis790  6674 Jun 10  2014 child_exp_dir_setup.pl -rwxrwxrwx. 1   2303 cis790  5233 Jun 10  2014 child_exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790   900 Jun 10  2014 convert.pl -rwxrwxr-x. 1 root   cis790  1912 Jun 10  2014 copySph.pl -rwxrwxr-x. 1 root   cis790  2152 Jun 10  2014 createTranscript.pl -rwxrwxr-x. 1 root   cis790   729 Jun 10  2014 createdict.pl -rwxrwxr-x. 1 root   cis790   182 Jun 10  2014 find.pl -rwxrwxr-x. 1 root   cis790    88 Jun 10  2014 genFileIDs.csh -rwxrwxr-x. 1 root   cis790   118 Jun 10  2014 genPhones.csh -rwxrwxr-x. 1 root   cis790   956 Jun 10  2014 lm_create.pl -rwxrwxr-x. 1 root   cis790  1199 Jun 10  2014 parseDecode.pl -rw-r--r--. 1 root   root    2658 Apr 27  2014 monoGen2.pl -rwxr-xr-x. 1   2310 cis790  2339 Apr 22  2014 extract23.pl -rwxr-xr-x. 1 root   root    1405 Apr 11  2014 buildData.pl -rw-r--r--. 1 root   root     159 Apr 10  2014 updateCFG.sh -rwxr-xr-x. 1 frs2000 cis790 9764 Apr  8  2014 fs_setup_SphinxTrain.pl -rwxr-xr-x. 1   2311 cis790   520 Mar 28  2014 trans_time.pl -rw-r--r--. 1   2311 cis790   959 Mar 26  2014 checkTrain.pl -rwxr-xr-x. 1   2310 cis790   915 Mar  4  2014 createSubTranscript.pl -rw-r--r--. 1   2310 cis790   123 Mar  3  2014 cleanTrans.sh -rw-r--r--. 1 root   root   55153 Feb 28  2014 buildme.sh -rw-r--r--. 1   2310 cis790   450 Feb 26  2014 pullFromTrans.pl -rw-r--r--. 1   2310 cis790   719 Feb 13  2014 gen_errors.pl -rwxr-xr-x. 1   2310 cis790  2765 Feb  6  2014 dictionary.pl -rwxr-xr-x. 1   2204 cis790 15377 Aug 31  2013 awesomeAdd.pl -rwxr-xr-x. 1   2204 cis790  5300 Aug  6  2013 clone_exp.pl 3/9: I just wanted to post my findings from going through the Wiki Scripts page and running some of the scripts that weren't documented that I could. I found that some of the issues with some of the scripts being documented on the scripts page and not being in the scripts/users directory was fixed. There were some other scripts that were in the scripts/users directory that weren't documented on the Wiki Scripts page. Findings from scripts/users directory regarding the fact that some of the scripts have the text in the directory that is green and there are just some words (like DELETE) that are colored blue. There are others that aren't colored. I'm thinking that the green ones are the scripts that are either the most recently used or the most rencently edited or created. The two that are in blue are directories (like DELETE). It looks like the ones that aren't colored are ones that either haven't been used recently or haven't been edited or created recently. It also looks as if some things have been moved or completely deleted. buildData2.pl (Created April 11 2014) Does the exact same function as the buildData.pl that is listed on the Scripts page under buildData. It links and copies all files related to the corpus data into the experiment. The user provides the corpus subset and data set as well as the experiment number. Based on these inputs, the script will link to the audio/wav, trans/train.trans, info/train.dic, info/train_train_fileids, info/train.phones and create the filters file. Note: trans/train.trans now points to the edited trans file. The unedited trans file is now trans/base.trans. checkTrain.pl (Created Mar 26 2014) This does the exact same function as the checkTrain.pl that is listed on the Scripts page under checkTrain. This script checks that each transcript entry matches a corresponding .sph file. Each .sph file has a 4 digit number to identify it. Each transcript entry has a 5 digit number (4 matching and 1 for multiples). This script simply grabs the 4 digits from each and makes sure they all have a match. It takes the name of the training directory as an argument. clone_exp.pl (Created Aug 6 2013) Does the exact same function as the clone_exp.pl that is listed on the Scripts page under clone_exp.pl. This script is designed to clone one experiment based on the settings from another. Allows for complete cloning of transcripts, dictionaries, phone lists, fileid lists, and wave files. This script is designed to clone an existing experiment. It will either clone the dictonaries and phone list; the transcripts, file list, and wavfiles; or it will do both. It will not touch the sphinx_train.cfg file or create feats from the copied wav files; use train_02.pl and make_feats.pl respectively to do those tasks. convert.pl (CJun 10 2014) Does the exact same function as the clone_exp.pl that is listed on the Scripts page under clone_exp.pl.This script will make symbolic links to all the required sph files that are noted in a transcript file located in a particular corpus directory. It first sets corpus directory, then appends the path the trans file based on the corpus directory provided, and then opens the transcript file for processing. copySph.pl (Created Jun 10 2014) There are copySph2.pl and copySph3.pl. I don't yet know what the difference is between these two but will find that out. createTranscript.pl (Created Jun 10 2014) For this page the first major thing that doesn’t involve coding that I could probably fix is that if you go to the Scripts page and then go createTranscript.pl there are two description sections and they could combine them into one. Besides this the script works similar to createTranscript.pl. The algorithm they used for that script was based on the same algorithm I used to calculate the total time as 308 hours (which we determined was not accurate due to channel overlap). This script builds a new transcript using the algorithm I used to calculate 250 hours. It also allows for the user to specify a start time, thus allowing us to build transcripts of data in the middle of our full transcript rather than being constrained to the start or end. This script will create a smaller transcript that is of a length of time specified by the user. The length_of_time is in seconds. The Start time indicates how far into the transcript the script should go before it starts to copy dialog to the new transcript. Time is also in seconds. createSubTranscript.pl (Created Mar 4 2014) Does this exact same thing as described in the createSubTranscript.pl on the Scripts page. This script creates a new transcript file, derived from the base transcript provided, of the specified number of hours and starting at the specified hour. This script uses the same time calculation as corpusSize2.pl, which differs from the way Sphinx calculates time (Sphinx does not account for overlap in the audio files). dictionary.pl (can't find in directory) Scripts log says this is the same as createdict.pl (Jun 10 2014) There's also dictionary2.pl and dictionary3.pl will explore more and see what they do and what the difference between them is find.pl (Created Jun 10 2014) Looks like this just searches for a term in the cmudict.0.6d ductionary specifically. The page for this script is really sparse. I don’t know if there needs to be anything more added to this. If so I could run it and try to do this. gen_errors.pl (Feb 13 2013) This script does the same thing that is described on the Scripts page under gen_errors.pl. This script is used when training the acoustic model for an experiment. It gives the output of how many errors were encountered and fills the html file with how many errors were found in each step.
 * Results:

The page for this script is really sparse. I don’t know if there needs to be anything more added to this. If so I could run it and try to do this. generateFeats.pl (Jul 15 2015) This script does the same thing that is described on the Scripts page under generateFeats.pl. This script calls scripts_pl/makeFeats.pl and removes the symbolic link to the audio files, and creates a new link pointing to the actual audio directory. Run this script after prepareExperiment.pl to fully prepare an experiment up to acoustic model training. genFileIDs.csh (Jun 10 2014) I couldn’t figure out how to run this script, so I’m going to guess that it just does the same as described on the Scripts page under genFileIDs. It says that it uses the sed command, which parses and transforms text. also know as a substitution command. genTrans.pl (Feb 3 3016) The Scripts log says there should be multiple versions for this file, but I can only find this one. There's no record that in the log that they were deleted or something. On the scripts page under what I think is the most recent version of this script there’s no description of what the script does.

The page could use some updating, so that it shows what the most recent script is and what the script does right on the page that describes the script. I could try to do this. lm_create.pl (Jun 10 2014) This script does exactly what is described on the scripts page under lm_create. It creates a Sphinx Language Model from a text file.

For the majority of the other scripts when you click on the name it will show some of the code for the script, but this one doesn’t. I could try to add some of the code to the page. master_run_train.pl (Feb 25 2014) In the scripts/user directory there's record of a master_run_train_old.pl. In the Scripts log there's no record of this file, so I will try and find out what this does and if it's different from the master_run_train.pl.

It looks like it got updated. Now in both the scripts/user directory and in the scripts log there is a record of master_run_train. This script is designed to help ease the tedious process of Running a Train. The base foundation of this script is to instead having the user call the script and passing a large number of arguments with hard-to-understand argument titles (i.e. -a, -e, -n, -x etc ...), the user just inputs the required arguments as input through the terminal. parseDecode.pl (Jun 10 2014) This file does exactly as described in the scripts log. It says that the script takes all the lines that begin with FWDVIT and dump them into a temp file. These are the predicted lines that we need to look at to compare to the original statement. Removes the parsed file if it exists. Opens the temp file. Reads and manipulates each line in the temp so that it matches the format that the transaction log is in and outputs the reformatted string. 7212014prepareExperiment.pl (Seems like this is the most recent one). I don't know why it was named the way it is when it is the only prepareExperiment file in the scripts/user directory. So maybe eventually rename this file.

This script isn’t in the scripts/users directory anymore. I don’t know why. It could have been purposefully moved or accidentally deleted. pruneDictionary (Mar 11 2015) This file does exactly as described in the scripts log. It says that this is a complete rebuild of the pruneDictionary script which significantly improves the performance of generating a dictionary file. What once could take hours on larger data now takes seconds

train_01.pl (Aug 6 2013) This script isn’t in the scripts/users directory anymore. I don’t know why. It could have been purposefully moved or accidentally deleted. train_02.pl (Aug 6 2013) This script isn’t in the scripts/users directory anymore. I don’t know why. It could have been purposefully moved or accidentally deleted. trans_time.pl (Mar 28 2014) This script does exactly as described in the scripts wiki. It takes a transcript file as an argument and prints the number of lines, and outputs the time in seconds and hours. updateDict It's listed in the Scripts log in the Wiki, but I can't find in scripts/user directory createWikiExperiment.pl (Jul 3 2015) This script does exactly as described on the scripts wiki page. It grabs the next available experiment number. A new wiki page is created for the experiment and a link to that page is added to the Speech:Exps list of experiments. There are a number of steps involved to make this script work. The script connects to the foss.unh.edu server and attempts to login with the given username and password. The scripts gets a login token and passes it back to the server to create a connection with the user logged in. The script then gets an edit token to allow the user to make changes to the wiki. Next the next free experiment number is received and stored. Afterwards the user enters the author's name and a brief description of the experiment they wish to create. The script then creates the new experiment entry on the Speech:Exps page. Finally the script creates the new experiment's own page with the information provided by the user. This script used the HTTP::Response, HTTP:Cookies, and LWP::UserAgent modules to function in addition to the information provided by the MediaWiki API createWiki_Sub_Experiment.pl (Feb 4 2015) This script does exactly as described on the scripts wiki page. It grabs the next available sub experiment number for a given experiment. A new wiki page is created for the sub experiment and a link to that page is added to the main experiment's page list of experiments. The user's password is visible when he or she enters it in. It is very similar to the createWikiExperiment script and follows most of the same process. There are some other Scripts that are in the scripts/user directory, but not documented or listed in the Scripts log on the Wiki page setup_SphinxTrain.pl (Feb 3 2016) Usage: To setup a new SphinxTrain task Create the new directory (e.g., mkdir RM1)

Go to the new directory (e.g., cd RM1)

Run this script (e.g., perl       $SPHINXTRAIN/scripts_pl/setup_SphinxTrain.pl -task RM1)

./scripts_pl/setup_SphinxTrain.pl -help For full list of arguments

./scripts_pl/setup_SphinxTrain.pl [-force] [-sphinxtraindir ] -task For setting up the SphinxTrain environment, located at , into current directory, naming the task

exp_dir_setup.pl (Jul 3 2015)

run_decode.pl (Apr 6 2015) Runs Sphinx 3 decoding job usage: run_decode.pl   

insertSil.pl (Mar 4 2015)

exp_sphinx_config.pl (Jan 18 2015)

addQueUsers.pl (Jun 10 2014)

child_exp_dir_setup.pl (Jun 10 2014)

child_exp_sphinx_config.pl (Jun 20 2014)

corpusSize.pl (Jun 10 2014)

genPhones.csh (Jun 10 2014)

monoGen2.pl (Apr 27 2014)

extract23.pl (Apr 22 2014)

updateCFG.sh (Apr 11 2014)

fs_setup_SphinxTrain.pl (Apr 10 2014) Usage: To setup a new SphinxTrain task Create the new directory (e.g., mkdir RM1)

Go to the new directory (e.g., cd RM1)

Run this script (e.g., perl       $SPHINXTRAIN/scripts_pl/setup_SphinxTrain.pl -task RM1)

./scripts_pl/setup_SphinxTrain.pl -help For full list of arguments

./scripts_pl/setup_SphinxTrain.pl [-force] [-sphinxtraindir ] -task For setting up the SphinxTrain environment, located at , into current directory, naming the task

corpusSize0.pl (Apr 6 2014) This script isn’t in the scripts/users directory anymore. I don’t know why. It could have been purposefully moved or accidentally deleted. Or it’s totally possible that this just wasn’t needed.

corpusSize2.pl (Mar 4 2014) This script isn’t in the scripts/users directory anymore. I don’t know why. It could have been purposefully moved or accidentally deleted. Or it’s totally possible that this just wasn’t needed.

cleanTrans.sh (Mar 3 2014)

buildme.sh (Feb 28 2014)

pullFromTrans.pl (Feb 26 2014)

awesomeAdd.pl (Aug 31 2013)

3/2: During class I will work on writing a brief description of what the scripts in the directory do, so that we can go through and update the Scripts page on the Wiki and make sure that each of the pages about the scripts is up to date. 3/7: Even though I haven't gotten a chance to work on writing descriptions of more of the scripts in the scripts/users directory I'm planning on doing more tomorrow and then Wednesday I will hopefully feel comfortable enough that I won't screw anything up by running any of the scripts that are in the directory but aren't documented. 3/8: Today I went through the rest of the scripts in the scripts/users and wrote descriptions of them. There were some that I found that haven't been documented on the Scripts Wiki page that I could easily get a description of what they are used for and the ones that I couldn't figure out what they do they gave me permission denied. Tomorrow I will see if I can find more information about the scripts in the scripts/users directory that aren't documented at all. 3/9: I just wanted to post my findings from going through the Wiki Scripts page and running some of the scripts that weren't documented that I could. I found that some of the issues with some of the scripts being documented on the scripts page and not being in the scripts/users directory was fixed. There were some other scripts that were in the scripts/users directory that weren't documented on the Wiki Scripts page. So today during class Matthew and I are going to go through my notes and start making the necessary fixes. 3/7: My only concern as of right now is that without there being any documentation on some of the scripts I don't know what code to start them and I don't want to screw something up. So I'm thinking that I might do some research to find out how to run them. 3/9: I have no concerns right now. I'm just glad I was successfully able to get my part of the work this week done.
 * Plan:
 * Concerns:

Week Ending March 22, 2016
3/11: During class on Wednesday I saw that there's a strange file listed in the scripts/users directory. I don't know I'm the only one that can see it. All I know is that I ran some files last week, but I didn't create anything. So I have no idea where the file came from. 3/23: I still need to get comfortable with running a train and decode, but right now I'm worried that I'm going to screw something up.
 * Task:

3/11: During class on Wednesday I saw that there's a strange file listed in the scripts/users directory. I don't know I'm the only one that can see it. All I know is that I ran some files last week, but I didn't create anything. So I have no idea where the file came from. -rw-r--r--. 1 mgm28  cis790     0 Mar  8 10:43 phone 3/23: The directory is looking better than it did last week there were 24 added. total 336 -rwxr-xr-x. 1 root   root    4107 Mar 23 13:48 makeTest.pl drwxr-xr-x. 24 root   root    4096 Mar 21 18:01 History -rwxr-xr-x. 1 root   cis790   243 Mar 21 17:59 scriptsPath.csh -rwxr-xr-x. 1 jrs1036 cis790 7377 Mar 20 21:09 genUttAudio.pl -rwxrwxr-x. 1 root   root    2598 Mar 18 16:34 makeCorpus.pl -rwxr-xr-x. 1 root   root     743 Mar  5 10:00 corpusSize.pl -rwxrwxrwx. 1 root   root     410 Mar  3 19:08 generateFeats.pl -rwxr-xr-x. 1   2412 cis790  6288 Feb  4 17:00 createWiki_Sub_Experiment.pl -rwxr-xr-x. 1 root   root    2997 Feb  3 16:12 prepareTrainExperiment.pl -rwxrwxrwx. 1   2418 cis790  4473 Feb  3 16:07 genTrans.pl -rwxrwxr-x. 1 root   cis790  9832 Feb  3 15:49 setup_SphinxTrain.pl -rwxr-xr-x. 1 root   root    2997 Jul 24  2015 prepareDecode.pl -rwxr-xr-x. 1   2303 cis790  4982 Jul  3  2015 exp_dir_setup.pl -rwxr-xr-x. 1 root   root    5467 Jul  3  2015 createWikiExperiment.pl -rwxr-xr-x. 1 root   root     762 Apr  6  2015 run_decode.pl -rwxr-xr-x. 1 root   root    1405 Mar 11  2015 pruneDictionary.pl -rwxrwxrwx. 1   2418 cis790   303 Mar  4  2015 insertSil.pl -rwxrwxrwx. 1   2405 cis790 30270 Feb 25  2015 master_run_train.pl -rwxrwxrwx. 1   2303 cis790  4368 Jan 18  2015 exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790  1003 Jun 10  2014 ParseTranscript.pl -rwxrwxr-x. 1 root   cis790  1716 Jun 10  2014 addQueueUsers.pl -rwxrwxrwx. 1   2303 cis790  6674 Jun 10  2014 child_exp_dir_setup.pl -rwxrwxrwx. 1   2303 cis790  5233 Jun 10  2014 child_exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790   900 Jun 10  2014 convert.pl -rwxrwxr-x. 1 root   cis790  1912 Jun 10  2014 copySph.pl -rwxrwxr-x. 1 root   cis790  2152 Jun 10  2014 createTranscript.pl -rwxrwxr-x. 1 root   cis790   729 Jun 10  2014 createdict.pl -rwxrwxr-x. 1 root   cis790   182 Jun 10  2014 find.pl -rwxrwxr-x. 1 root   cis790    88 Jun 10  2014 genFileIDs.csh -rwxrwxr-x. 1 root   cis790   118 Jun 10  2014 genPhones.csh -rwxrwxr-x. 1 root   cis790   956 Jun 10  2014 lm_create.pl -rwxrwxr-x. 1 root   cis790  1199 Jun 10  2014 parseDecode.pl -rw-r--r--. 1 root   root    2658 Apr 27  2014 monoGen2.pl -rwxr-xr-x. 1   2310 cis790  2339 Apr 22  2014 extract23.pl -rwxr-xr-x. 1 root   root    1405 Apr 11  2014 buildData.pl -rw-r--r--. 1 root   root     159 Apr 10  2014 updateCFG.sh -rwxr-xr-x. 1 frs2000 cis790 9764 Apr  8  2014 fs_setup_SphinxTrain.pl -rwxr-xr-x. 1   2311 cis790   520 Mar 28  2014 trans_time.pl -rw-r--r--. 1   2311 cis790   959 Mar 26  2014 checkTrain.pl -rwxr-xr-x. 1   2310 cis790   915 Mar  4  2014 createSubTranscript.pl -rw-r--r--. 1   2310 cis790   123 Mar  3  2014 cleanTrans.sh -rw-r--r--. 1 root   root   55153 Feb 28  2014 buildme.sh -rw-r--r--. 1   2310 cis790   450 Feb 26  2014 pullFromTrans.pl -rw-r--r--. 1   2310 cis790   719 Feb 13  2014 gen_errors.pl -rwxr-xr-x. 1   2310 cis790  2765 Feb  6  2014 dictionary.pl -rwxr-xr-x. 1   2204 cis790 15377 Aug 31  2013 awesomeAdd.pl -rwxr-xr-x. 1   2204 cis790  5300 Aug  6  2013 clone_exp.pl
 * Results:

3/11: I don't know what the plan for the week is.
 * Plan:

3/11: I don't know what the plan for the week is, like if we are supposed to still be contacting and communicating with the members of the little groups that we are a part of to figure out what our tasks for the week are?
 * Concerns:

Week Ending March 29, 2016
3/23: The task for this week will probably be to implement an idea for organization with the scripts/user directory so that there is some sort of organization. My idea for the organization of the scripts/user directory is by semester because it had the dates for when it was created or when it was last edited and a new version was updated. It would probably benefit me to run another train and run a decode. 3/28: I did more going through the file that I took notes in and cleaned that up, so that it has descriptions of what each of the scripts in the scripts/user directory does. 3/30: I did more going through the file that I took notes in and cleaned that up, so that it has descriptions of what each of the scripts in the scripts/user directory does.
 * Task:

3/23: total 336 -rwxr-xr-x. 1 root   root    4107 Mar 23 13:48 makeTest.pl drwxr-xr-x. 24 root   root    4096 Mar 21 18:01 History -rwxr-xr-x. 1 root   cis790   243 Mar 21 17:59 scriptsPath.csh -rwxr-xr-x. 1 jrs1036 cis790 7377 Mar 20 21:09 genUttAudio.pl -rwxrwxr-x. 1 root   root    2598 Mar 18 16:34 makeCorpus.pl -rwxr-xr-x. 1 root   root     743 Mar  5 10:00 corpusSize.pl -rwxrwxrwx. 1 root   root     410 Mar  3 19:08 generateFeats.pl -rwxr-xr-x. 1   2412 cis790  6288 Feb  4 17:00 createWiki_Sub_Experiment.pl -rwxr-xr-x. 1 root   root    2997 Feb  3 16:12 prepareTrainExperiment.pl -rwxrwxrwx. 1   2418 cis790  4473 Feb  3 16:07 genTrans.pl -rwxrwxr-x. 1 root   cis790  9832 Feb  3 15:49 setup_SphinxTrain.pl -rwxr-xr-x. 1 root   root    2997 Jul 24  2015 prepareDecode.pl -rwxr-xr-x. 1   2303 cis790  4982 Jul  3  2015 exp_dir_setup.pl -rwxr-xr-x. 1 root   root    5467 Jul  3  2015 createWikiExperiment.pl -rwxr-xr-x. 1 root   root     762 Apr  6  2015 run_decode.pl -rwxr-xr-x. 1 root   root    1405 Mar 11  2015 pruneDictionary.pl -rwxrwxrwx. 1   2418 cis790   303 Mar  4  2015 insertSil.pl -rwxrwxrwx. 1   2405 cis790 30270 Feb 25  2015 master_run_train.pl -rwxrwxrwx. 1   2303 cis790  4368 Jan 18  2015 exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790  1003 Jun 10  2014 ParseTranscript.pl -rwxrwxr-x. 1 root   cis790  1716 Jun 10  2014 addQueueUsers.pl -rwxrwxrwx. 1   2303 cis790  6674 Jun 10  2014 child_exp_dir_setup.pl -rwxrwxrwx. 1   2303 cis790  5233 Jun 10  2014 child_exp_sphinx_config.pl -rwxrwxr-x. 1 root   cis790   900 Jun 10  2014 convert.pl -rwxrwxr-x. 1 root   cis790  1912 Jun 10  2014 copySph.pl -rwxrwxr-x. 1 root   cis790  2152 Jun 10  2014 createTranscript.pl -rwxrwxr-x. 1 root   cis790   729 Jun 10  2014 createdict.pl -rwxrwxr-x. 1 root   cis790   182 Jun 10  2014 find.pl -rwxrwxr-x. 1 root   cis790    88 Jun 10  2014 genFileIDs.csh -rwxrwxr-x. 1 root   cis790   118 Jun 10  2014 genPhones.csh -rwxrwxr-x. 1 root   cis790   956 Jun 10  2014 lm_create.pl -rwxrwxr-x. 1 root   cis790  1199 Jun 10  2014 parseDecode.pl -rw-r--r--. 1 root   root    2658 Apr 27  2014 monoGen2.pl -rwxr-xr-x. 1   2310 cis790  2339 Apr 22  2014 extract23.pl -rwxr-xr-x. 1 root   root    1405 Apr 11  2014 buildData.pl -rw-r--r--. 1 root   root     159 Apr 10  2014 updateCFG.sh -rwxr-xr-x. 1 frs2000 cis790 9764 Apr  8  2014 fs_setup_SphinxTrain.pl -rwxr-xr-x. 1   2311 cis790   520 Mar 28  2014 trans_time.pl -rw-r--r--. 1   2311 cis790   959 Mar 26  2014 checkTrain.pl -rwxr-xr-x. 1   2310 cis790   915 Mar  4  2014 createSubTranscript.pl -rw-r--r--. 1   2310 cis790   123 Mar  3  2014 cleanTrans.sh -rw-r--r--. 1 root   root   55153 Feb 28  2014 buildme.sh -rw-r--r--. 1   2310 cis790   450 Feb 26  2014 pullFromTrans.pl -rw-r--r--. 1   2310 cis790   719 Feb 13  2014 gen_errors.pl -rwxr-xr-x. 1   2310 cis790  2765 Feb  6  2014 dictionary.pl -rwxr-xr-x. 1   2204 cis790 15377 Aug 31  2013 awesomeAdd.pl -rwxr-xr-x. 1   2204 cis790  5300 Aug  6  2013 clone_exp.pl makeTest.pl (March 26) Script produced by Peter with assistance from Matthew Heyner Usage: makeTest.pl [-d] 0287/001 0271/003 300hr For the flag... -d means use dev.trans -e means use eval.trans -t means use train.trans Please make sure you have ran a train in the source (and if you are not using a flag, make sure you have created a language model). If this has not happened, then this code will not work properly.
 * Results:

prepareTrainExperiment.pl (Feb 3) usage (run from main experiment directory): prepareExperiment  : The name of the corpus and data set. Ex: 10hr/train

prepareDecode.pl (July 24 2015) usage (run from main experiment directory): prepareExperiment  : The name of the corpus and data set. Ex: 10hr/train

generateFeats.pl (Jul 15 2015) Description:This script calls scripts_pl/makeFeats.pl and removes the symbolic link to the audio files, and creates a new link pointing to the actual audio directory. Run this script after prepareExperiment.pl to fully prepare an experiment up to acoustic model training. Code:
 * 1) ! /usr/local/bin/perl

use Cwd; use File::Basename;

$path = getcwd; $exp = basename($path); $cmd = "scripts_pl/make_feats.pl -ctl etc/$exp". "_train.fileids"; system($cmd); $cmd = "rm wav"; system($cmd); $cmd = "ln -s /mnt/main/corpus/switchboard/full/audio/utt wav"; system($cmd); print "\nComplete!\nRun \"nohup scripts_pl/RunAll.pl. &\" to begin training.\n";

genTrans.pl (Feb 3 3016) usage: genTrans.pl    Example: tiny/train 0011_train should be executed from the top level experiment directory ex: /mnt/main/Exp/001

master_run_train.pl (Feb 25 2014)

=
===================================== Script Title: Training Master Script Version: 2.5 Date Modified: 2/25/2014 Purpose: This script is designed to make the process of Running a Train faster and more efficient as this process is one that ecompasses a lot of uses. Please follow the instructions for each step and carefully read which arguments its asking for.

=
=====================================

=
======== PART 1 ===================== Before we begin, we must get some simple information. Terms we might ask: MASTER Experiment example: /mtn/main/Exp/0200 CHILD Experiment example: /mnt/main/Exp/0200/d12/s2000

Is this a new MASTER or CHILD Experiment? (type 'm' for MASTER or 'c' for CHILD) m

Your Experiment Number is going to be: 273

Before you continue, please go the Experiment Log and add your entry for the 273 Experiment. A guide has been made to walk you through this by going here: http://foss.unh.edu/projects/index.php/Speech:Add_Experiment_Entry_Guide

It's okay ... we'll be right here waiting for you to compelte that ...

If you want to cancel and NOT have a new Experiment Directory to be created, please press control + C

Once you have completed the new entry, please type '1' to continue: 1

=
======== PART 2 ===================== Tasks: - Create the new Experiment directory on Caesar. - Prep the new Experiment directory with needed sub-folders and essential Perl scripts.

Using Experiment Number: 273 Executing Part 1 - perl /mnt/main/scripts/user/exp_dir_setup.pl -a 1 -x 273

Output from script: Successfully created 0273 experiment directory. Please type '1' to continue: 1

=
======== End Part 2 =====================

=
======== PART 3 ===================== Tasks: - Modify the sphinx_train.cfg Configuration file. - Lines 6, 7, 8 we have to specify our Experiment number, our base Experiment directory path, and the base Experiment directory path on Casear. - Lines 79 and 80 we have to switch around which line is commented out and which one isn't. We want to use the line: CFG_HMM_TYPE = '.cont.' - Line 107 we can specify a custom value for the Density. - Line 120 we can specify a custom value for the Senone setting.

Arguments needed: OPTIONAL - Please enter a value to be used as the Density (8, 16, 32, 64 are commonly used): 8 OPTIONAL - Please enter a value to be used as the Senone (default is 1000): 1^H8 Using Desnity value: 8 Using Senone value: 8 Executing Part 3 - perl /mnt/main/scripts/user/exp_sphinx_config.pl -x 273 -d 8 -s 8

Output from script: Error: Cannot open config file! at /mnt/main/scripts/user/exp_sphinx_config.pl line 89. Now opening.../mnt/main/Exp/273/etc/sphinx_train.cfgPlease type '1' to continue: 1

=
======== End Part 3 =====================

=
======== PART 4 ===================== Tasks: - Generate the Transcripts we will use for this Train.

pruneDictionary (Mar 11 2015) Usage: pruneDictionary4.pl

createWiki_Sub_Experiment.pl (Feb 4 2015) Usage: Grabs the next available sub experiment number for a given experiment. A new wiki page is created for the sub experiment and a link to that page is added to the main experiment's page list of experiments. The user's password is visible when he or she enters it in. It is very similar to the createWikiExperiment script and follows most of the same process.

setup_SphinxTrain.pl (Feb 3 2016) Usage: To setup a new SphinxTrain task Create the new directory (e.g., mkdir RM1)

Go to the new directory (e.g., cd RM1)

Run this script (e.g., perl        $SPHINXTRAIN/scripts_pl/setup_SphinxTrain.pl -task RM1)

./scripts_pl/setup_SphinxTrain.pl -help For full list of arguments

./scripts_pl/setup_SphinxTrain.pl [-force] [-sphinxtraindir ] -task For setting up the SphinxTrain environment, located at , into current directory, naming the task

run_decode.pl (Apr 6 2015) Runs Sphinx 3 decoding job usage: run_decode.pl <taskName> <trainingExpID> <SenoneCount>

convert.pl (June 10 2014) usage: copySph.pl <corpus_dir> <corpus_dir> Example: /mnt/main/corpus/switchboard/mini/train

copySph.pl (June 10 2014) usage: copySph.pl <corpus_dir> <corpus_dir> Example: /mnt/main/corpus/switchboard/mini/train

createTranscript.pl (June 10 2014) usage: createTranscript.pl <input_transcript> <output_transcript> <length_of_time> <start_time> Duration is in seconds

find.pl (June 10 2014) Usage find.pl

lm_create.pl (June 10 2014) Creates a Sphinx Language Model from a text file usage: lm_create.pl inFile

parseDecode.pl(June 10 2014) usage: parseDecode.pl

extract23.pl (April 22 2014) usage: genTrans.pl <corpus_dir> <exp_id> <corpus_dir> Example: tiny/train 0011_train should be executed from the etc level experiment directory ex: /mnt/main/Exp/0011/etc

buildData.pl(April 11 2014) usage (run from main experiment directory): buildData.pl <corpus_dir> <exp_id> <corpus_dir>: The name of the corpus and data set. Ex: 10hr/train

fs_setup_SphinxTrain.pl(April 8 2014) Usage: To setup a new SphinxTrain task Create the new directory (e.g., mkdir RM1)

Go to the new directory (e.g., cd RM1)

Run this script (e.g., perl        $SPHINXTRAIN/scripts_pl/setup_SphinxTrain.pl -task RM1)

./scripts_pl/setup_SphinxTrain.pl -help For full list of arguments

./scripts_pl/setup_SphinxTrain.pl [-force] [-sphinxtraindir <SphinxTrain directory>] -task For setting up the SphinxTrain environment, located at <SphinxTrain directory>, into current directory, naming the task

createSubTranscript.pl (March 4 2014) corpusSize.pl <corpus_base_name> <total_hours> <offset_hours>

dictionary.pl (Feb 6 2014) Compares the list of words in a file to the words in a dictionary and outputs the words available with pronunciations. This script is intended to be used in conjunction with pruneDictionary2.pl        Usage: dictionary2.pl <WordFile> <DictionaryFile> <OutputFile>

3/23: The plan for this week will probably be to implement an idea for organization with the scripts/user directory so that there is some sort of organization. My idea for the organization of the scripts/user directory is by semester because it had the dates for when it was created or when it was last edited and a new version was updated. It would probably benefit me to run another train and run a decode. 3/28: I did more going through the file that I took notes in and cleaned that up, so that it has descriptions of what each of the scripts in the scripts/user directory does. Tomorrow I'll probably go through more of the files and write descriptions for what they do. Wednesday morning (making up for the work that I couldn't do on Sunday, due to being away all day travelling to and from a funeral) I will finish writing description for the rest of the scripts in the directory and record them in my log as well as send it in an email to my group members. 3/30: I did more going through the file that I took notes in and cleaned that up, so that it has descriptions of what each of the scripts in the scripts/user directory does. 3/23: I don't really have any concerns as of right now. 3/28: The only concern that I have right now is what description I should put for the files that I can't run because I get a permission denied message.
 * Plan:
 * Concerns:

Week Ending April 5, 2016
4/2: My tasks for this week consist of running another train. This time I'll also also Create the Language Model and run the Decode in addition to running the train. I'm also going to do some research for my team. 4/5: I was successful in getting some good accurate research done for my team and recorded. This evening I'm planning on running a train, creating the language model, and run the decode. 4/6: I was going to meet with Saverna before class, but that didn't happen, so hopefully during class we will meet.
 * Task:

4/5: I was successful in getting some good accurate research done for my team and recorded. 4/6: I was going to meet with Saverna before class, but that didn't happen, so hopefully during class we will meet.
 * Results:

4/2: The plan for this week is running another train and also Create the Language Model and run the Decode. I'm also going to do some research for my team. 4/5: I was successful in getting some good accurate research done for my team and recorded. This evening I'm planning on running a train, creating the language model, and run the decode 4/6: I was going to meet with Saverna before class, but that didn't happen, so hopefully during class we will meet. 4/2: My only concern as of right now is I remember during class on Wednesday when Saverna was running her Train, Creating the Language Model, and running the Decode I know you have to add old to the path name and I don't remember which ones they were. To solve this concern I sent an email to my group. 4/5: Right now I don't have any concerns. 4/6: Right now I don't have any concerns
 * Plan:
 * Concerns:

Week Ending April 12, 2016
4/7: I don't yet know what I'm supposed to be doing this week, so to solve this I have emailed Jon to find out. I'm hoping I can try to either run another whole train and the other two processes or possibly go into the folder that contains the only train I have ever run and create the Language Model and Decode it. 4/10: I now know that the next step is to do more research to develop our "secret sauce". 4/12: I have been doing some research on one of the parameters to see if we can get a better result by changing it. 4/13: I have been doing some research on one of the parameters to see if we can get a better result by changing it.
 * Task:
 * Results:

4/10: I now know that the next step is to do more research to develop our "secret sauce". 4/12: I have been doing some research on one of the parameters to see if we can get a better result by changing it. 4/7: My only concern as of right now is that I don't yet know what I'm supposed to be doing this week, so to solve this I have emailed Jon to find out. 4/10: I don't have any concerns. 4/12: My only concern right now is that I'm just hoping I can find some information to be of usefulness to my group.
 * Plan:
 * Concerns:

Week Ending April 19, 2016

 * Note: I was in NYC this weekend and had no time on Friday, Saturday, or Sunday due to being busy and exhausted. Then Monday was spent catching up on sleep and doing some research.

4/13: I have been doing some research on one of the parameters to see what other values we can change it to. Also finding out if we can get a better result by changing it. 4/19: I have been doing some more research on one of the parameters to see what other values we can change it to and if we change it if we can get a better result.
 * Task:

4/19: I have found a lot about the parameter that I'm interested in, but nothing about what values it could be changed to.
 * Results:

4/13: My plan for this week is researching a parameter and seeing what else it could be changed to in order to get a better result.
 * Plan:

4/13: As of right now I have no concerns. 4/19: As of right now the only concern that I have is the fact that I have found a lot about the parameter that I'm interested in, but nothing about what values it could be changed to
 * Concerns:

Week Ending April 26, 2016

 * Note: There wasn't much more detail I could put without revealing more to the other group.

4/25: I went back to look at the power point that we (the Captain America team) started going over before the URC to try to make sense of everything.
 * Task:

4/25: I went back to look at the power point that we (the Captain America team) started going over before the URC to try to make sense of everything and now it's starting to make sense to me.
 * Results:

4/25: The plan is to keep doing research to see how we can best get the lowest WER percentage.
 * Plan:

4/25: I have been doing some research the past couple of days and have yet to find anything helpful. So far what I have found is changing values that we already did and it made the WER go up.
 * Concerns:

Week Ending May 3, 2016
4/29: I don't yet know what I'm supposed to be doing this week.
 * Task:


 * Results:

4/29: I don't yet know what I'm supposed to be doing this week, but it probably is more research to see what paramaters we can change to get the best result possible. 4/29: My concern right now is that I don't know what I'm supposed to be doing. Now that we have combined groups I don't know who to email to find out what I should be doing, so I sent a quick email off to Jon to find out.
 * Plan:
 * Concerns:

Week Ending May 10, 2016
5/5: My task for this week is going through the document that my group (the experiment group) created about what we have done as tasks this semester and fixing the grammar because there are some things that just aren't worded right and lots of commas are missing. 5/7: My task for today will be finishing up editing the document about what tasks we have done because there's some of it that just doesn't read right. 5/11: I successfully finished editing the document detailing what we did for the Experiment Group.
 * Task:

5/11: I successfully finished editing the document detailing what we did for the Experiment Group.
 * Results:

5/5: My task for today is going through the document that my group (the experiment group) created about what we have done as tasks this semester and fixing the grammar because there are some things that just aren't worded right and lots of commas are missing. For the rest of the week I'm hoping to make some good contributions to the final document that we as a class or that we in our two split groups have to submit. 5/7: My task for today will be finishing up editing the document about what tasks we have done because there's some of it that just doesn't read right. 5/11: I successfully finished editing the document detailing what we did for the Experiment Group.
 * Plan:

5/5: As of right now I don't have any concerns. 5/7: As of right now I still have no concerns.
 * Concerns: