Quick Run
This page quickly guides you through the process of installing the latest version of ALO and installing the sample AI Solution (Titanic) to register an AI Solution without modifying the code. If you can run Linux commands and have an environment with Jupyter notebook installed, you can proceed with the following steps. (Note: Some commands may differ if Jupyter notebook is running inside a docker container.)
Topics
step1: Start ALO
For more details, please refer to the installation page (Install ALO).
step1-1: Install ALO
Write the desired working folder name in {project_name}.
git clone https://github.com/mellerikat/alo.git {project_name}
cd {project_name}
step1-2: Set up ALO runtime environment
Below is an example of setting up an Anaconda virtual environment, but you can set up any virtual environment that can run Python 3.10, such as pyenv.
conda init bash ## Initialize conda environment
exec bash ## Restart bash
# conda create --prefix /home/jovyan/testenv python=3.10 ## Method to install a non-deletable virtual environment when running jupyter in docker
conda create -n {virtual_env_name} python=3.10 ## 3.10 is required
conda activate {virtual_env_name} ## If conda activate does not work, run exec bash
pip install -r requirements.txt
## [Additional work for running the solution registration Jupyter Notebook]
pip install ipykernel
## If the virtual environment name is /home/jovyan/testenv, write only testenv.
python -m ipykernel install --user --name {virtual_env_name} --display-name {jupyter-virtual-env-display-name}
Note : For setting up a virtual environment using pyenv + pipenv, please refer to the Install ALO - Development Environment Setup page.
step2: Develop Solution
If you need to understand the AI Solutions running in ALO, you can check the sample solution 'titanic' by installing it.
step2-1: Install and run Titanic
# Install the solution folder in the {project_home} location where ALO main.py exists
git clone https://github.com/mellerikat/titanic.git solution
python main.py # Run ALO
step2-2: Install and run AI Contents
AI Contents refers to reusable AI Solutions designed to solve various industry challenges using AI. These can be applied to tasks by simply changing data in the YAML file settings without additional development. For more details, refer to the AI Contents page (AI Contents).
Note: For Git code access to AI Contents, refer to (AI Contents Access).
| AI Contents | Description |
|---|---|
| TCR | TCR is AI Contents for classification/regression on tabular data, automatically performing preprocessing, sampling, HPO, and optimal model selection. https://github.com/mellerikat-aicontents/Tabular-Classification-Regression.git |
| GCR | GCR has the same input/output as TCR, but uses Graph ML technology to extract and utilize hidden information from the data, improving model accuracy and handling data with missing values. https://github.com/mellerikat-aicontents/Graph-powered-Classification-Regression.git |
| FCST | FCST is AI Contents that analyzes time series data and predicts future values. https://github.com/mellerikat-aicontents/Forecasting.git |
| CV | CV is AI Contents for automating the classification of various types of classes through image classification models. https://github.com/mellerikat-aicontents/Vision-Classification.git |
| PAD | PAD stands for Point Anomaly Detection and is AI Contents for detecting whether point-type time series data is normal or abnormal. https://github.com/mellerikat-aicontents/Anomaly-Detection.git |
| TAD | TAD stands for Tabular AnomalyDetection, which is AI Contents that detects normal and abnormalities through univariate or multivariate variables in the form of a table. https://github.com/mellerikat-aicontents/Tabular-Anomaly-Detection.git |
| MAD | MAD is AI Contents used for early sensing of anomalies or defects by monitoring multiple variables simultaneously and detecting anomalies. http://mod.lge.com/hub/dxadvtech/aicontents/mad.git |
| VAD | Visual Anomaly Detection (VAD) is an AI content that detects abnormal images by learning normal images. https://github.com/mellerikat-aicontents/Vision-Anomaly-Detection.git |
Running AI Contents is the same as running the Titanic sample solution. Just refer to the Git addresses in the table above and change them accordingly.
# Login Caching
git config --global credential.helper 'cache --timeout=864000'
# Install the solution folder in the {project_home} location where ALO main.py exists
git clone {AI Contents git address} solution
python main.py # Run ALO
step3: Register Solution
To register an AI Solution, follow the procedures in the register-ai-solution.ipynb guide.
- Set up Infra for solution registration
- Write AI Solution information
- Execute solution registration
- Conduct AI training test
step3-1: Set up Infra
Write the environment information for registering the AI Solution in ./{project_home}/setting/infra_setup.yaml. For more details, refer to the Infra setup page (Configure Infra).
"VERSION": 1.0 ## solution_metadata version
"AIC_URI": "https://aicond.meerkat-ai.com/"
"REGION": "ap-northeast-2"
"WORKSPACE_NAME": "meerkat-ws"
"BUILD_METHOD": "codebuild" ## codebuild, docker, buildah
"LOGIN_MODE": "static" ## ldap, static
"REPOSITORY_TAGS":
- "Key=Owner,Value=Meerkat"
"AWS_KEY_PROFILE": "alo-aws-profile"
### aws codebuild (cloud build)
## 'type': 'WINDOWS_CONTAINER'|'LINUX_CONTAINER'|'LINUX_GPU_CONTAINER'|'ARM_CONTAINER'|'WINDOWS_SERVER_2019_CONTAINER'|'LINUX_LAMBDA_CONTAINER'|'ARM_LAMBDA_CONTAINER',
"CODEBUILD_ENV_TYPE": "LINUX_CONTAINER"
## 'computeType': 'BUILD_GENERAL1_SMALL'|'BUILD_GENERAL1_MEDIUM'|'BUILD_GENERAL1_LARGE'|'BUILD_GENERAL1_XLARGE'|'BUILD_GENERAL1_2XLARGE'|'BUILD_LAMBDA_1GB'|'BUILD_LAMBDA_2GB'|'BUILD_LAMBDA_4GB'|'BUILD_LAMBDA_8GB'|'BUILD_LAMBDA_10GB',
"CODEBUILD_ENV_COMPUTE_TYPE": "BUILD_GENERAL1_SMALL"
step3-2: Write AI Solution Information
Run the login cell in the register-ai-solution.ipynb Jupyter Notebook using the assigned account, then fill in and execute the following cell content. The key field is 'solution_name', and the name of the AI Solution to be registered should be entered in lowercase letters and hyphens. For more details, refer to the AI Solution registration page (Register AI Solution).
#----------------------------------------#
# Write AI Solution Spec #
#----------------------------------------#
solution_info ={
'solution_name': 'my-solution',
'inference_only': False,
'solution_update': False,
'overview': "Write an overall description of the AI Solution.",
'detail': [
{'title': 'title001', 'content': 'content001'},
{'title': 'title002', 'content': 'content002'},
],
'contents_type': {
'support_labeling': False,
'inference_result_datatype': 'table', # image
'train_datatype': 'table', # image
'labeling_column_name': ''
},
'train_gpu': False,
'inference_gpu': False,
"inference_arm": False
}
step3-3: Execute AI Solution Registration
Execute the following steps in register-ai-solution.ipynb.
import sys
try:
del sys.modules['src.solution_register'], sys.modules['src.constants']
except:
pass
from src.solution_register import SolutionRegister
## set register instance
infra = "./setting/infra_config.yaml"
register = SolutionRegister(infra_setup=infra, solution_info=solution_info)
## login AI Conductor
register.login(username, password)
## register AI solution
register.run()