In this section, you will be introduced to some of the concepts that are frequently refered in the API and documentation.
Project is a unit "workspace" that contains all the files and configurations pertaining to your machine learning API. Each project has a special file named
roro.yml file is a configuration file used to describe the project and it's runtime environment.
- Project name
- Runtime environment
- Services to deploy
- Periodic taks
face-detection/ |___ predict.py |___ requirements.txt |___ roro.yml |___ train.py
The structure for
roro.yml file is given below:
project: <your-project-name> runtime: <runtime-choice> services: - name: default function: <file-name>.<function-name> size: <hardware-size> # optional - name: <name> function: <file-name>.<function-name> size: <hardware-size> # optional tasks - name: <task-name> command: <python command> size: <hardware-size> # optional when: <specify time frequency>
# roro.yml project: face-detection runtime: python3-keras services: - name: default function: predict.predict size: M1 tasks: - name: train command: python train.py size: G1 when: every day at 11:55
Each project has a specific
runtime environment, which defines the base software environment used for the project. The
runtime includes two docker images, one for CPU and another one for GPU.
runtime environment for each project is specified in the
roro.yml configuration file. List of available runtimes are given below
|python3||Python 3.X with Scikit Learn, Pandas, and other common Pydata libraries|
|python3-tensorflow||python3 + Tensorflow|
|python3-keras||python3 + Tensorflow + Keras|
|python3-pytorch||python3 + PyTorch|
|python2||Python 2.X with Scikit Learn, Pandas, and other common Pydata libraries|
|python2-keras||python3 + Tensorflow + Keras|
You can change the runtime of your project any time through the dashboard.
Help us improve the documentation. Flag errors, issues or request how-tos, guides and tutorials on our
#documentation channel on our Slack