Speech:Spring 2017 Jeremy Beal Log


 * Home
 * Semesters
 * Spring 2017
 * Proposal
 * Report
 * Information - General Project Information
 * Experiments - List of speech experiments

Week Ending February 7th, 2017

 * Task:
 * 2/2/17 - Check on work from previous semester(s).
 * 2/4/17 - Research some of the software already set up for the project from last semester.
 * 2/7/17 - Check on progress of others in the current class.
 * 2/8/17 - Download and configure UNH VPN.


 * Results:
 * 2/2/17 - Some of the software the previous semester worked on was already installed, and they either did not need to update it, or updated it, in addition to looking for new additions.
 * 2/4/17 - Looked over the list at Speech Software Functionality. Learned some of the functions of CMU.
 * 2/7/17 - Need to take care of getting the VPN set up so I can work outside of class.
 * 2/8/17 - Installed the VPN with Pulse Secure. I was able to SSH into Caesar from my home desktop.


 * Plan:
 * 2/2/17 - I need to look more into what has been done from last semester.
 * 2/4/17 - I need to learn more about the work from the previous semester, and where the system is currently at in work.
 * 2/7/17 - Download and configure UNH VPN.
 * 2/8/17 - Meeting with team members today, we will need to discuss what to work on next.


 * Concerns:
 * 2/2/17 - Need to figure out what work we need to do during the semester, compared to last semester.
 * 2/4/17 - There is going to be a lot of research to do, and we will need to find out how each piece of software interacts with others.
 * 2/7/17 - I still need to learn how we will go about making changes to the systems.
 * 2/8/17 - I currently do not have the information written down on how to connect to the other systems.

Week Ending February 14, 2017

 * Task:
 * 2/8-2/14 - Unfortunately, I was sick this week, and was not able to get any work done.


 * Results:
 * 2/8-2/14 - Unfortunately, I was sick this week, and was not able to get any work done.


 * Plan:
 * 2/8-2/14 - I will work hard to catch up with the other members of my team.
 * Concerns:
 * 2/8-2/14 - I will work hard to catch up with the other members of my team. I do not want to be far behind.

Week Ending February 21, 2017

 * Task:
 * 2/17 - Since I missed the previous class, I took time to read the logs of my other team members to see what I had missed. I also checked over our group's proposal again.
 * 2/20 - I need to read through the logs from the previous semester's Tools group, as well as my team mates and some of the other groups to get an idea of where the project is at now.
 * 2/21 - I am going to spend my time today researching the differences between GCC, which should already be installed on the systems, and G++, which we are looking to install.
 * 2/22 - I will research the differences between Sphinx 3 and PocketSphinx, and what installation for PocketSphinx may require or change.


 * Results:
 * 2/17 - Our plan is to further research information on G++ and PocketSphinx. We updated our proposal to reflect these changes. Other groups may also need to work with us in the future for installing other software. We will be testing with Majestix, once it can be made usable.
 * 2/20 - The tools group was still working on researching the software for the systems around this time last year. They focused on recent editions of the software already on the system, before looking to add anything. This differs with our plans to look at G++ and PocketSphinx, before checking on the current software. It looks like Majestix is working, so we will be able to test with that once we have enough information to work with. We seem to all have a better understanding of G++ and what changes it may lead to, but still need to look more at PocketSphinx once that is in good shape.
 * 2/21 - G++ seems easy to install, so most of the work for getting it set up on Majestix will be checking what files may change with its installation.
 * 2/22 - Sphinx 3 and PocketSphinx have the same installed dependencies - gcc, automake, autoconf, libtool, bison, swig at least version 2.0, python development package, and pulseaudio development package. PocketSphinx requires the sphinxbase to be installed, which is a part of Sphinx 3. Sphinx 3 may be better for larger sample sizes in testing, but PocketSphinx is said to be the fastest version, and is also based in C like Sphinx 3 was.


 * Plan:
 * 2/17 - I need to spend more time researching the benefits and changes of G++ and PocketSphinx, compared to GCC and Sphinx 3 respectively. We also need to keep up to date with the Systems group on the status of Majestix. After we take care of G++ and PocketSphinx, we will look at the software already installed to check for any changes or updates we may need to make.
 * 2/20 - Need to look more into the installation of PocketSphinx and what files it may change when we try to install, what software may not work with it, etc. I plan to do that in the next few days.
 * 2/21 - Before we install G++, we will need to run a train to get a comparison for after we install it. Our team will be researching both G++ and PocketSphinx, so I will look into both to get an idea of the work that will be needed for each. This week, we plan to run the train for pre-G++.
 * 2/22 - Once we are able to install PocketSphinx on Majestix, we will want to run tests on varying sample sizes to check the differences in speed, accuracy, etc. between Sphinx 3 and PocketSphinx. We also will need to check what files it may change compared to the installation of Sphinx 3, of course.


 * Concerns:
 * 2/17 - Unfortunately I missed out on a week's worth of work, so I will need to work to catch up with my team on our research for our work.
 * 2/20 - I have been a bit behind still on my research compared to my group. I am planning on working to open up more time outside of class for me to work on this course over the semester, hopefully in time for next week.
 * 2/21 - It sounds like we need to get G++ working as quickly as we can, so that other groups can make use of it, so we will need to be efficient in getting it all set up.
 * 2/22 - I still have not personally run a test train, and will need to do so with my team or on my own soon.

Week Ending February 28, 2017

 * Task:
 * 2/25/17 - We need to research more on how we may be able to install Majestix and PocketSphinx on Majestix, before we can do it on Caesar. Currently Majestix is offline, unlike Caesar, so we either need to connect Majestix to the Internet or to find an offline way of installing software.
 * 2/26/17 - Checked on the logs for my team members to see how their week has been going.
 * 2/28/17 - Read logs from others in the class to see where progress is at.


 * Results:
 * 2/25/17 - There are a couple ways of installing Ubuntu packages offline, using other programs. Keryx is a free, open source program to put software onto a USB, which can then be used to install it on the offline machine. There is also the Synaptic package download script, which generates a script on the offline machine to install the software, uses the script from a USB on an online machine to grab the package, then can install it on the offline machine from the USB. These both seem like valid options for offline installs for Majestix.
 * 2/26/17 - It sounds like no one has tried to install G++ yet, but we're getting close to it. Since we are about done with that, I will continue looking into the installation of PocketSphinx.
 * 2/28/17 - Our group is still having some issues getting Majestix to successfully run a test train. We also are still having issues getting Majestix to unmount from Caesar, and to find ways of installing G++ remotely. I talked with a couple members of the Systems group, and it sounds like we won't be able to get Majestix an Internet connection at the moment, so we will need to try out one of the offline installation methods. I am still looking into more possible ways to do this, and hope to have something in time for class.


 * Plan:
 * 2/25/17 - I will look into getting one or both of the mentioned running on one of my computers if I can. I may need to get a Ubuntu VN working in order to do it.
 * 2/26/17 - If no one does it before Wednesday, we can work on installing G++ onto Majestix this week while we are all together. Between now and then, I have more research to do for installing software and PocketSphinx.
 * 2/28/17 - Look further into offline installation of software for Majestix. Consult with my group and others int he class on the issues we seem to be having with Majestix, so we can run a full train before going through with the G++ installation. We should check with other groups that are using the other blades to see if they have run into any similar issues.
 * Concerns:
 * 2/25/17 - In order for Keryx or Synaptic to work, we will need to hook a monitor up to Majestix, and we will need to work with the systems group to access it. We also are still having some odd issues when trying to unmount Majestix, which may cause issues in the future.
 * 2/26/17 - If we aren't able to get Keryx or Synaptic to work, we will need to get Internet access for Majestix; the systems group should know more about how we can do it, but it may put our group's schedule back.
 * 2/28/17 - As mentioned with my plans, we need to find a way to install onto Majestix while it is offline, and to run a train on it.

Week Ending March 7, 2017

 * Task:
 * 3/1 - Work with the group to successfully decode and score a train on Majestix so we can move on with the G++ installation.


 * 3/6 - Look through Majestix and the logs from the previous semester to look for the snapshot files that they may have made before and after their software installations.


 * 3/7 - Checked logs for my team's progress this week.


 * 3/8 - Checked logs for team's progress.


 * Results:
 * 3/1 - We were unable to get Majestix to run the decode and train, but were able to do so through Caesar. It seems there are functions missing form Majestix from when it was fixed earlier this semester, so we will be handing it back to the Systems group, and instead start working with Obelix. Before we do so, we will need to look for the snapshot files from the previous semester, if they were not deleted when Majestix was fixed.


 * 3/6 - I was unable to find any of the files in Majestix. Daisuke's logs mention a snapshot directory, with subdirectories for before and after each installation, but I was unable to find the snapshot folder. I looked through each directory in Majestix, but found nothing that may have been the last semester's work. I checked with my team members, and they were also unable to find any of the log files. It seems that they were wiped out with the machine.


 * 3/7 - We are starting to run trains on Obelix, now that we have finished with Majestix. Jeff copied the files that he could find from Majestix over to Caesar, in case we need those.


 * 3/8 - We look to be ready to move ahead with installing GCC and G++ today.


 * Plan:
 * 3/1 - I need to look through Majestix's directories to try and find the files. The logs from the previous semester will likely mention where they were saved to, so we can start there.
 * 3/6 - Since we cannot find the files, the Systems group will take back Majestix, and we can start working with Obelix once they say it is ready.


 * 3/7 - Now that we are up and running with Obelix, hopefully we can get the snapshots done on Wednesday, and move on to the installation of G++.


 * 3/8 - I wil work with our group to get GCC installed today, and possibly G++ if we have the time.


 * Concerns:
 * 3/1 - Our schedule for getting G++ working is getting pushed back with the continued issues with Majestix; hopefully we can get Obelix ready this week.

previous Tools group.
 * 3/6 - I am worried that we may be missing much of the documentation from last year due to not having the files, so we will have to rely on the logs from the


 * 3/7 - The groups will be splitting in to teams this week; hopefully we can still all work on Obelix to get through the G++ and PocketSphinx installations.


 * 3/8 -None currently.

Week Ending March 21, 2017

 * Task:
 * 3/8 - We are looking to install GCC and/or G++ today.
 * 3/20 - Read logs to see how my team (and the Rebels) are doing.
 * 3/21 - Attempt to snapshot Obelix's files post-GCC and pre-G++.
 * 3/22 - Read logs to see how my team (and the Rebels) are doing.


 * Results:
 * 3/8 - We were able to successfully install GCC onto Obelix. Before we can install G++, we will need to run a train/decode/score on Obelix, and snapshot the files beforehand. Hopefully we can get this done over break.
 * 3/20 - Sharayah plans on running a train on Obelix, so we will be able to use this data to further our process of installing G++ onto Obelix.
 * 3/21 - I had issues logging into Obelix using my own credentials (possible ssh key issue?), so I had to use root. Using the directories listed by the Tools group last semester, I created a new directory in Obelix, /snapshot and /snapshot/gcc/post-gcc, and copied the files that they had to there from Obelix's current state. I also created /snapshot/g++/pre-g++ and put them there.
 * 3/22 - It sounds like we may be ready to try and install G++ today. I will work with our group to try and get this done.


 * Plan:
 * 3/8 - I will work with my team to get a train going on Obelix so we can see how it stands with just GCC, but not G++. I will also look into snapshotting the files from Obelix to have a mark of where we currently are.
 * 3/20 - I will plan on trying to snapshot Obelix to get a mark of what the files look like post-GCC and pre-G++.
 * 3/21 - With the snapshots set up, if we can run a full train/decode/score on Obelix, we should be in line for installing G++.
 * 3/22 - Depending on how class goes today, I can try to run a train on Obelix tomorrow to get more data for us to use.
 * Concerns:
 * 3/8 - None at this time.
 * 3/20 - None at this time.
 * 3/21 - None at this time.
 * 3/22 - None at this time.

Week Ending March 28, 2017

 * Task:
 * 3/22-2/28 - Unfortunately, I was sick again this week, and was not able to get much work done. I did stay in contact with my group to see what I had missed, and what we were working on next.


 * Results:
 * 3/22-2/28 - Unfortunately, I was sick again this week, and was not able to get much work done.


 * Plan:
 * 3/22-2/28 - Unfortunately, I was sick again this week, and was not able to get much work done.
 * Concerns:
 * 3/22-2/28 - None at this time.

Week Ending April 11, 2017

 * Task:
 * 4/5 - Try to install Miniconda on Miraculix and work on the URC poster.
 * 4/10 - Read logs.
 * 4/11 - Read logs.
 * 4/12 - Look in to how to hotfix G++ on Caesar for the drones.


 * Results:
 * 4/5 - There were issues with the Miniconda installation, so that will be worked on another day. Sharaya and I were able to put together our group's URC poster. We also gave information to installing G++ onto Caesar, and are waiting or approval to install it. I did more research into the information for potentially installing PocketSphinx.
 * 4/10 - Read logs. The URC poster has been submitted.
 * 4/11 - Read logs.
 * 4/12 - We have not had to hotfix any installations yet, but the 2016 Tools group documented how they did it for EMACS. We can discuss it when we talk with Jonas about the G++ installation and how we will go about doing it (https://foss.unh.edu/projects/index.php/Speech:Spring_2016_Jonathan_Trimble_Log from April 30th). After we do the install, we'll do snapshots to make sure everything went correctly.


 * Plan:
 * 4/5 - Finish the Miniconda installation. Fix any errors or lack of information on the URC poster before we submit it. Work on the G++ installation and PocketSphinx proposal.
 * 4/10 - Read logs. We should be able to finish the miniconda installation tomorrow.
 * 4/11 - Read logs. We need to research the possibility of installing G++ onto Caesar as a hotfix so that the drones can use it.
 * 4/12 - Install G++ on Caesar and hotfix it for use on the drones.


 * Concerns:
 * 4/5 - We are short on time left in the semester if we want to install PocketSphinx. Otherwise none.
 * 4/10 - None at this time.
 * 4/11 - None at this time.
 * 4/12 - None at this time.

Week Ending April 18, 2017

 * Task:
 * 4/16/17 - Checking in and reading logs.
 * 4/17/17 - Check with Systems group for progress on Caesar backup. Confirm process of creating a hotfix for the GCC installation.
 * 4/18/17 - Checking in and reading logs.
 * 4/19/17 - Add to the PocketSphinx report. Confirm information on gcc and g++ installations and prepare for possibly installing today.


 * Results:
 * 4/16/17 - Checking in and reading logs.
 * 4/17/17 - Caesar's main directory has been backed up, so pending approval from Jonas we should be good to go forward with the GCC installation on Caesar this week. We may need to confirm with him on Wednesday for this. Like previously stated on my logs last week, we have a description of how to create the hotfix installation files on Caesar for use with the drones; the steps are listed below (edited from Jonathan Trimbles' originals for GCC).

*Create a link to gcc executable in /usr/bin (while in /usr/bin): ln -s /mnt/main/local/lib/hotfix/gcc/usr/bin/emacs-23.1-nox gcc

*Link gcc hotfix libexec dir to current server's /usr/libexec dir named gcc (while in /usr/libexec): ln -s /mnt/main/local/lib/hotfix/gcc/libexec/gcc gcc

*Link gcc hotfix usr/share dir to current server's /usr/share dir named gcc (while in /usr/share): ln -s /mnt/main/local/lib/hotfix/gcc/usr/share/gcc gcc

Of course, these directories and files will be created once we actually start the installation process, and will be added to the installation doc made for it.
 * 4/18/17 - Checking in and reading logs.
 * 4/19/17 - I added more details to the PocketSphinx report. I confirmed the directories affected by our gcc and g++ installations on Obelix and how we will undergo the installation for Caesar.


 * Plan:
 * 4/16/17 - From what I can gather, we are still waiting on the systems group to finish backing up Caesar to move forward with the GCC and G++ installation on it. I'll check in with them tomorrow and see what their progress is on that, and then we can see about doing the installation on Wednesday before the URC.
 * 4/17/17 - Check to install GCC (and possibly G++) this week. Update/add to PokcetSphinx installation proposal.
 * 4/18/17 - Checking in and reading logs. I will work on the PocketSphinx proposal later today before class.
 * 4/19/17 - We hopefully can start the gcc and g++ installations today; we may just install on Caesar today, and work on the hotfixes for the drones over the week.


 * Concerns:
 * 4/16/17 - None at this time.
 * 4/17/17 - None at this time.
 * 4/18/17 - None at this time.
 * 4/19/17 - None at this time.

Week Ending April 25, 2017

 * Task:
 * 4/23/17 - Checking in and reading logs.
 * 4/24/17 - Checking in and reading logs.
 * 4/25/17 - Check to see if we have heard back from Jonas yet. Check with team mates on any possible progress we can make in the meantime.


 * Results:
 * 4/23/17 - Checking in and reading logs.
 * 4/24/17 - Checking in and reading logs.
 * 4/25/17 - We still have not heard back to see if we can move ahead with the GCC installation. For now, we grabbed some pre-GCC snapshots on Miraculix, but have not installed it yet because we were told not to work on the machine previously.


 * Plan:
 * 4/23/17 - We are all waiting to hear back from Jonas before continuing the GCC/G++ install. I will start getting ready to add GCC and G++ to the software table already on the wiki that was created last semester.
 * 4/24/17 - Still waiting to hear back from Jonas.
 * 4/25/17 - We will need to check with Jonas in class tomorrow if we do not hear back from him by then.


 * Concerns:
 * 4/23/17 - None at this time.
 * 4/24/17 - None at this time.
 * 4/25/17 - We are running short on time but unable to progress further on our plans.

Week Ending May 2, 2017

 * Task:
 * 5/2/17 - Checking in and reading logs.
 * 5/3/17 - Checking in and reading logs.


 * Results:
 * 5/2/17 - Checking in and reading logs.
 * 5/3/17 - Checking in and reading logs.


 * Plan:
 * 5/2/17 - Checking in and reading logs.
 * 5/3/17 - Checking in and reading logs.
 * Concerns:
 * 5/2/17 - None at this time.
 * 5/3/17 - None at this time.

Week Ending May 9, 2017

 * Task:
 * 5/7/17 - Checking in and reading logs.
 * 5/8/17 - Checking in and reading logs.
 * 5/9/17 - Go over and edit the team's final report and the group's section for class report.
 * 5/10/17 - Work on group's portion of final class report and documentation.


 * Results:
 * 5/7/17 - Checking in and reading logs.
 * 5/8/17 - Checking in and reading logs.
 * 5/9/17 - Helped with the final revisions for the team's final report. Also worked on the group report.
 * 5/10/17 - Worked on group's portion of final class report and documentation.


 * Plan:
 * 5/7/17 - Checking in and reading logs.
 * 5/8/17 - Checking in and reading logs.
 * 5/9/17 - Do a final check over group and team documents.
 * 5/10/17 - Help with final class report.


 * Concerns:
 * 5/7/17 - None at this time.
 * 5/8/17 - None at this time.
 * 5/9/17 - None.
 * 5/10/17 - None.