Installing AI Contents

Updated 2024.05.05

Install domain-specific AI Contents to solve various industry challenges.

Topics

• Running AI Contents

---

Running AI Contents

Set up a Python 3.10 virtual environment in a Linux environment beforehand. Explore AI Contents and install the AI Contents suitable for your problem-solving needs. Install the contents of the AI Contents Git into the solution folder located in the same path as ALO's main.py.

cd {solution_name}

# Cache login credentials
git config --global credential.helper 'cache --timeout=864000'

git clone {git_url} solution

All AI Contents are structured as follows:

./{solution_name}/solution
    └ experimental_plan.yaml # Mandatory
    └ experimental_plan_extra.yaml # Optional
    └ sample_data/  # Madatory
        └ train/
             └ train_sample1.csv
             └ train_sample2.csv
             └ train_sample3.csv
        └ test/
             └ test_sample.csv
    └ guide/ # Optional
        └ modeling_guide.ipynb

The experimental_plan.yaml is a configuration file that contains information related to the setup of the AI Solution. It allows the construction of the training and inference ML pipeline, modification of key hyperparameters for data and modeling, and execution of the ML pipeline.
You can use the sample data present in the sample_data folder as input to run the AI Contents. The subfolders under sample_data are separated into training and inference sample data, and the folder names can be anything other than train and test.
Once the experimental_plan.yaml is completed, you can run the AI Contents wrapped with ALO by executing the main.py command as shown below. When ALO runs, it automatically installs the dependent Python modules written in the asset requirements for each asset source.

python main.py

---