Write UI Parameter
In the experimental_plan.yaml file, you can set parameters that users can modify in the EdgeConductor UI. This allows the AI Solution developer to predefine essential parameters required for the project and clearly specify which parameters the AI Solution user can change.
By setting up these parameters in the configuration file, users can easily adjust and experiment with the model by simply entering or selecting the necessary parameter values.
The experimental_plan.yaml supports limited UI parameter types, such as Float, Integer, String, Single Selection, and Multi Selection, to enhance usability, reduce the potential for errors, and prevent users from mistakenly setting critical parameters incorrectly.
Once the AI model has been thoroughly tested and its performance verified in the development environment of the AI Solution, it can be registered and deployed directly to the operational stage on the Mellerikat platform using the register-ai-solution.ipynb Jupyter notebook.
The UI parameters set by the AI Solution developer will also be automatically reflected in the EdgeConductor UI and shown to the user.
Topics
Declaring UI Parameters
Write the parameters provided as a UI in Edge Conductor in experimental_plan.yaml. First, select the items to be displayed in the UI from user_parameters and write the parameter names in ui_args.
#experimental_plan.yaml
user_parameters:
- train_pipeline:
- step: input
args:
- input_path: train
x_columns: [input_x0, input_x1, input_x2, input_x3]
y_column: label
ui_args:
- x_columns
- y_column
Write detailed information for the parameters listed in ui_args in the ui_args_detail section of experimental_plan.yaml.
Detailed Description of UI Parameters
#experimental_plan.yaml
ui_args_detail:
- train_pipeline:
- step: input
args:
- name: x_columns
description: TCR 모델링에 사용될 x columns를 ','로 구분하여 기입합니다. ex) x_column1, x_column2
type: string
default: ''
range:
- 1
- 1000000
- name: y_column
description: TCR 모델링에 사용될 y column명을 기입합니다. ex) y_column
type: string
default: ''
range:
- 1
- 1000000
You can choose from float, int, string (or string type list), single_selection, and multi_selection types for the type of UI args. Write the required information for each type in the ui_args_detail section.
Note: Each AI Contents manual page lists the available UI Parameters and detailed information that can be used.
-
float or int
- Supports int and float types.
- The range indicates the min and max value range.
- name: float_param
description: this is param to input float
type: float
default:
- 0.5
range:
- 0.0
- 1.0 -
string
- Use when a string value needs to be entered.
- The range indicates the character length limit.
- name: string_param
description: this is param to input string
type: string
default:
- "hello"
range:
- 5
- 100-
To support a list, use the string type.
- If the user enters multiple values separated by commas (", ") in the string type parameter in the UI, the values are split and delivered as a list during training in AI Conductor. (e.g., "value1, value2, value3" --> ["value1", "value2", "value3"])
- Warning : If the user enters values like 1, 2, 3 in the string type parameter in the EdgeConductor UI, the entered values are delivered to ALO as ["1", "2", "3"]. Therefore, if the values in the list need to be integers instead of strings, type conversion code should be written inside the Asset.
- name: string_param
description: this is param to input string
type: string
default:
- "x1, x2, x3,"
range:
- 5
- 100
-
single_selection
- Use when one value needs to be selected from selectable values.
- name: single_selection_param
description: this is param to input single selection
type: single_selection
selectable:
- option1
- option2
- option3
default:
- option1 -
multi_selection
- Use when multiple values need to be selected from selectable values.
- name: multi_selection_param
description: this is param to input multi selection
type: multi_selection
selectable:
- option1
- option2
- option3
default:
- option1
- option3