Speech:Spring 2013 Justin Mulholland Log


 * Home
 * Semesters
 * Spring 2013
 * Proposal
 * Report

Week Ending February 5th, 2013
- Take a look at how the system is currently set up and try and see how it works. - Read Experiment Logs from previous Semesters. - Look at Unix commands and start becoming familiar with Perl.
 * Task:
 * Results:

2/2/13
 * Researched some Perl and UNIX material.

2/3/13
 * Continued Research. Read some previous experiment logs.

- SSH into Ceaser. - Research UNIX commands and Perl.
 * Plan:

- Learning Perl so quickly and using it effectively. - Remembering UNIX commands, other than the Basic commands.
 * Concerns:

Week Ending February 12, 2013
- try and catch up with Jake about our current status and what we should be focusing on. - get into Caesar and try and get more familiar with it.
 * Task:


 * Results:

2/6/13
 * Got into Caesar and navigated around the file system and the different directories.
 * Continued researching PERL and UNIX scripts

2/9/13
 * Got in touch with Jake, he informed me about his situation. Planned on setting up a skype session.
 * Read some more previous semester logs and information

2/10/13
 * Read others logs

2/12/13
 * Read Logs

- Get on track with what we need to accomplish. Continue to research PERL.
 * Plan:


 * Concerns:

Week Ending February 19, 2013

 * Task:

2/16/13
 * Results:
 * Still not entirely clear on the specific tasks Jake and I will be working on in regards to this project.
 * We got in contact with Prof. Jonas to set up a time where we can sit down and figure everything out.
 * Continued to look at PERL.

2/17/13
 * Spent some time looking at previous semesters and looked at more of the work that was accomplished.
 * Noticed Caesar was back up and running so I played around in there just looking at different directories and navigating around in there.

2/18/13
 * Read through other's logs

2/19/13
 * Read Logs

- Meet up with Prof. Jonas and get a clearer understanding of what Jake and I need to work towards accomplishing for our part of the project.
 * Plan:

- Getting to far behind and not getting as much work done as we would like.
 * Concerns:

Week Ending February 26, 2013
- Get on board with Jake and catch up on what I missed from Weds class (had to stay late at work). - Hold a meeting with Prof. Jonas to get some direction of where Jake and I should be heading in terms of our roles as the experiment group.
 * Task:

2/21/13
 * Results:
 * Jake and I had our meeting with Prof. Jonas to gain some clarity and direction towards our role as the experiment group for the project. This really helped us out as we now have a plan to work towards.
 * Sat down with Jake and set our plan for what we want to accomplish for our march 6 project installment, and also what we want to work on for the week.

2/22/2013
 * Looked at the logs of Cedric to try and get an understanding of what he accomplished during his time on this project, and how he went about trying to run a train. Since we will be working closely with the modeling group, I want to get a good understanding of how the process of running and decoding trains work. Provided Jake with some input for the proposal and After some editing and communication, we submitted it.

2/24/2013
 * Read others logs. Eric gives some really good information about what he has been working on in the modeling group, and this will help keep me and Jake up to date on what the group has been working on.

2/25/13
 * Read others logs.

- Hopefully start coding what we want to try accomplish. Collaborate with the modeling group. - Time..
 * Plan:
 * Concerns:

Week Ending March 5, 2013

 * Task:


 * Results:

3/4/13
 * Read others logs.

3/5/13
 * Read logs.


 * Plan:


 * Concerns:

Week Ending March 12, 2013
- Get back on track and start working on a script for the modeling group to assist them in creating/copying experiments more efficiently.
 * Task:

3/9/13
 * Results:
 * Read some logs of the other students.

3/10/13
 * Read others logs.


 * Plan:


 * Concerns:

Week Ending March 26, 2013
- Read over the sections of how to set up and run a train. Meet up with my group to go over the process. Continue to work with Jake on writing a script that will condense the steps/process for setting up an experiment and running a train.
 * Task:


 * Results:

3/20/13
 * Got caught up with Jake and discussed how we plan to continue working on the script for condensing the way experiments are set up.
 * We looked into using Perl but we are going to use Shell scripting. It could be done in either one, but shell seems like it will be easier.
 * Read through Cedric's log on how to create and run an experiment. We plan on breaking up the  Experiment Setup and Running the Train section into two different scripts for automating the whole process, that way it will simplify the whole process for setting up experiments.

3/24/13
 * Noticed that Eric updated the section on how to set up a train. Read through the document to get a feel for what it entails. He did a really nice job of explaining the process step by step.
 * I did see a potential error in one of the comments. It read "Do execute this in the root experiment folder! (/mnt/main/Exp", and I assumed he meant "Do NOT execute this in the root experiment folder!".
 * I made a note and informed him about it, because I wasn't 100% sure that's what he meant, and I didn't want to change it myself just in case.
 * Made my experiment directory. My number is 0033.
 * Re-formatted my logs so they are more consistent.

3/25/13
 * Performed the process of running a train. Generating the dictionary did not take as long as I thought it would (appx. 30 mins). Eric did a nice job of documenting the processs and made it really easy to follow.
 * Something came up and I didn't have enough time to actually run the train. Everything is set up though to the point right before running the train, so that's all I have left to do. I will be running it tomorrow afternoon.

3/26/13
 * Tried running a mini train but was not successful. I had to add a bunch of words to my dictionary, and three of them were not registering for some reason. It's getting late so I will address it with the group tomorrow to see if I can get it resolved.
 * I noticed a few guys were adding there experiments to the experiments page and I will also be adding an entry with a list of all the words I added.
 * Read others logs.

- Run a train on Monday (3/25) - Actually run the train tomorrow(3/26) and see what happens. Hopefully get the train running tommorow
 * Plan:
 * Concerns:
 * Thanks for pointing the typo out, its a good thing you caught that before somebody actually followed it! It was indeed a missing word. I did some formatting modifications earlier and I guess the 'not' got removed. I fixed it now so we should be good. Thanks again. -Eric Beikman

Week Ending April 2, 2013
- Get a train to run successfully as well as decode and score it.
 * Task:

3/27/13
 * Results:
 * Worked on trying to get the mini/train to run. Jake and Drew had the same errors when trying to run mini and tiny.
 * We plan on meeting tomorrow to start from scratch. I'm going to try and run the mini/eval, and hopefully that will be successful.

3/28/13
 * I documented my first experiment Mini Train Test and indicated the words I added as well as the error message I got.
 * Our group met up and restarted the process of running the trains.
 * Eric was with us as well and we informed him about the errors we were getting yesterday. He will be looking onto it.
 * Was Successful in running a mini/eval.
 * Mike was also successful with his mini eval train.
 * I documented the experiment Here

4/1/13
 * Built the language model. There were no problems there.
 * Ran the Decode and score process. The Decode ran fine but the score gave me the two errors that were noted in the instructions.
 * I documented my results Here

4/2/13
 * Read other students logs for the week.


 * Plan:
 * Get the next step in the process and figure out what we will be doing next.


 * Concerns:

Week Ending April 9, 2013

 * Task:


 * Figure out what the next step is.
 * Set up a new corpus and Run the first 5 hrs of the transcripts


 * Results:

4/3/13
 * Jake and I tried setting up the new corpus for group A&D while some of the others were getting caught up with running and decoding their trains.
 * We had trouble getting it to work, from setting up the transcript file to grabbing the copySph.pl file. There were permission issues, the transcript wasn't getting generated, etc.
 * I noticed Matt had started his own first_5hr corpus which will be for the A&D group, so I deleted our corpus and let him take control of the creating the new corpus.

4/4/13
 * Read other student's logs.

4/7/13
 * Read other logs to see if there was any progress made with setting up the new corpi.
 * As of today most logs are blank. The only progress that seems to be made is from Eric. I didn't see anything in Matt's log, but I did see Mike noted in his log that Matt was able to get the corpus set up with the audio files and everything, according to Tyler's Log.
 * I logged into Caesar and did see the audio files were there for the first_5hr corpus.
 * Not sure what the next step is, so I'll be contacting Drew to see if he knows what is going on and if there is a particular role I should be playing.

4/9/13
 * Read others logs to see if there was any progress with running trains. It looks like Eric and his team got the ball rolling with their train and started creating the acoustic model. Hopefully tomorrow Teams A&D can sit down, get together and work on getting our train running.


 * Plan:
 * (4/7) Talk to Drew and see if he know what the next step is.
 * Concerns:

Week Ending April 16, 2013
- Stay in the loop with the entire group (AD)and discuss what we will be working on and what needs to get done.
 * Task:

4/10/13
 * Results:
 * Received email from Charlie indicating he had talked to Matt about what he was working on script wise. Matt created his own script (genTrans2 or 3).
 * This sparked a chain email including everyone in the group.

4/13/13
 * Staying in the loop of the group email, we tried organizing a time where everyone can meet and run a train together. It's proving to be slightly difficult due to everyone's schedules.
 * Read Logs of other students.

4/14/13
 * Joined a Goggle hangout with most of Group AD (Tyler,Tom,Matt,Drew,Mike,Bego). Tyler shared his screen with his Putty Terminal so the group could see what he was doing. We ran into a few issues in the beginning, mainly with the genTrans script. The regular expressions Matt worked on had to be tweaked.
 * I had to leave the hangout at 11:00pm and by that point, the train was still not working. Hopefully the guys will get it working. Will keep up to date with Tyler tomorrow.

4/15/13
 * Received confirmation from Tyler that he and Matt were able to get the first_5hr train to finally run after, hours of working on it. (Thanks guys!)
 * He informed the group the language model was created and the decode was running. It is still running as of now (8:25pm).
 * Read logs of others.


 * Plan:


 * Concerns:

Week Ending April 23, 2013
- Get the poster done and run the first_5hr train and try and get a lower score.
 * Task:


 * Results:

4/17/13
 * Worked on the URC poster with the modeling group. We got most of it done in class.

4/20/13
 * Read logs of other students.

4/21/13
 * Tyler sent out an email to the entire group with a Google doc link that has the missing dictionary words. We were all asked to help out and add the correct pronunciations (phonetic spelling) to some of the words if we can. I added 16 phonetic spellings.

4/22/13
 * Read logs of other students.
 * Noticed that Caesar was not up.


 * Plan:


 * Concerns:

Week Ending April 30, 2013
- Work on the report. I am in charge of the scoring section.
 * Task:

4/27/13
 * Results:
 * Read Logs. Not much going on up to this point.
 * Looked over the report of Spring 2012 and 2011's Semester to get an idea of how the report will be written.

4/28/13
 * Read Logs of other students.

4/30/13
 * Wrote my section of the report. I'll most likely be adding some more info before the final draft is due.


 * Plan:


 * Concerns:

Week Ending May 7, 2013

 * Task:


 * Results:

5/6/13
 * Read logs

5/7/13
 * Read logs


 * Plan:


 * Concerns: