# Installation Notes for Development

## Running Eloquent App in Docker
The easiest way to install and run the Eloquent app is using the docker container from HuggingFace. You will need to have docker installed on your system (see prerequisites below) and then run the app container:

```
docker run -it -p 7860:7860 --platform=linux/amd64 \
	registry.hf.space/vgotcheva-eloquent:latest 
```

When the app container is fully started in docker, the app can be accessed in the browser with [http://localhost:7860](http://localhost:7860).

Eloquent App is a gradio app written in python. It uses ```automatic-speech-recognition``` and ```text-generation``` models through HuggingFace transformers library. 

The app has been tested on a laptop with specifications:
- CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
- RAM: 16.0 GB
- GPU: none
- SSD: at least 25GB available space

Note that the AI models (about 6GB total) will be downloaded and cached locally during the docker container start up. Therefore, it may take 10-15min for the app to start depending on the download speed. 

## Running Eloquent App in Python Environment
The app can be run in a python development environment. Please, check the prerequisites below. 

Clone the repository and cd in the cloned folder:
```
git clone https://huggingface.co/spaces/vgotcheva/eloquent
cd eloquent
```

Install the app requirements:
```
pip install --no-cache-dir --upgrade -r /code/requirements.txt
```

Start the app:
```
python app.py
```

Use the app by opening in browser [http://localhost:7860](http://localhost:7860).

## Prerequisites for development under Windows 11

### Visual Studio Code
https://code.visualstudio.com/

### Git for Windows including Git Credential Manager
https://gitforwindows.org/

### WSL 2 (with Git for Windows)
https://learn.microsoft.com/en-us/windows/wsl/install

### Linux Distribution in WSL
```
Ubuntu 20.04.6 LTS
```

### git and gif-lft
```bash
apt-get install git git-lfs
```

### Anaconda in Ubuntu
```bash
$ conda -V
conda 23.11.0
```

### conda virtual environment and pip
- https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#activating-an-environment
- https://anaconda.org/anaconda/pip

### Set up Git Credentials Manager in Ubuntu git
https://github.com/git-ecosystem/git-credential-manager/blob/release/docs/wsl.md 

```
git config --global credential.helper '/mnt/c/Program Files/Git/mingw64/bin/git-credential-manager.exe'
```
or add it directly to 
```
$ cat ~/.gitconfig
[user]
        name = demo
        email = demo@local
[credential]
        helper = /mnt/c/Program\\ Files/Git/mingw64/bin/git-credential-manager.exe
```

https://docs.gitlab.com/ee/ci/ci_cd_for_external_repos/

### Docker Desktop WSL 2 backend on Windows 
https://docs.docker.com/desktop/wsl/