Setup and Quickstart

[farinamhz]

Hey @arfaamr,

This is an issue page to log your progress in setting up and running the project.
Feel free to reach out if you have any concerns or questions.

@hosseinfani

[farinamhz]

@3ripleM.
Hey Medi,
If you provide a little description of dockerization here for Arfaa, as we discussed, that would be great!

[3ripleM]

Hi,
I've updated the readme of the project already. @arfaamr could you please run the container based on that?

docker run -it --name lady_container ghcr.io/fani-lab/lady:main

you just have to run above command. but before that make sure you have installed Docker engine on your computer.

[arfaamr]

Hi,
I have installed Docker and run the command, and can see the container for it in the Docker desktop app. I'm not familiar with Docker so it may take me some time to understand exactly what's happening.
I am away this weekend, but will continue with the setup next week.
Thank you for your help.

[arfaamr]

Hi, I'm not really sure where to go next. I tried following the next few steps listed in the README. It lists ./src/cmn, etc. to install some libraries. I assumed I should run these in the LADy container's shell, is that correct? It gives "Permission denied" errors, but I'm not sure why. Could I get some guidance?

[3ripleM]

Hi @arfaamr,

Actually, those are some directories, and you cannot execute directories like that. By running the container, you are using the full image of the LADy project, where all dependencies (libraries) are already installed in the Docker image. You can now proceed to the next step (Quickstart).

The reason for using Docker is reproducibility. To gain a deeper understanding of how an image is created, you can read the Dockerfile, where the setup process has been documented. Docker can be seen as an isolated container with its own files and libraries inside it, and it cannot use libraries from the machine on which it is running. Although the new version has some bugs, I'm working on fixing them. Some machines cannot run it at the moment, but if you are using the previous version of the Dockerfile, you are good.

In addition, I wanted to let you know that I spoke with Dr. Fani a while ago, and our new tasks involve "running" two libraries (BERT-E2E-ABSA and CAT) used in this project. We will be working on aspect-based sentiment analysis in the future. I have some suggestions for the running step, where I will create a new issue page, and we can discuss them there.

[AkiraShougun]

Hello,

I finished setting up the project and tried to run,

python main.py -naspects 5 -am rnd -data ../data/raw/semeval/toy.2016SB5/ABSA16_Restaurants_Train_SB1_v2.xml -output ../output/toy.2016SB5/

However I get this error:

[hosseinfani]

@AkiraShougun
that's a bug that we have not fix it yet due to the change to results' file structure. you can bypass it by removing the agg command in this line:

LADy/src/params.py

Line 10 in c261acb

'cmd': ['prep', 'train', 'test', 'eval', 'agg'], # steps of pipeline, ['prep', 'train', 'test', 'eval', 'agg']

Now, the next step is to trace the pipeline (debug and step line by line) and make sense out of the commands 'prep', 'train', 'test', 'eval'

Sometime next week, during office hours I can explain the high level explanation.