Flagging

Gradio provides several flagging callback classes to handle user feedback and data logging.

FlaggingCallback

class FlaggingCallback(ABC)

An abstract class for defining the methods that any FlaggingCallback should have.

Methods

setup

setup(components, flagging_dir)

This method should be overridden and ensure that everything is set up correctly for flag(). This method gets called once at the beginning of the Interface.launch() method.

components · Sequence[Component] · required

Set of components that will provide flagged data.

flagging_dir · str · required

A string containing the path to the directory where the flagging file should be stored.

flag

flag(flag_data, flag_option=None, username=None)

This method should be overridden by the FlaggingCallback subclass. This gets called every time the flag button is pressed.

flag_data · list[Any] · required

The data to be flagged.

flag_option · str | None · default: "None"

In the case that flagging_options are provided, the flag option that is being used.

username · str | None · default: "None"

The username of the user that is flagging the data, if logged in.

returns · int

The total number of samples that have been flagged.

SimpleCSVLogger

gr.SimpleCSVLogger()

A simplified implementation of the FlaggingCallback abstract class. Each flagged sample (both the input and output data) is logged to a CSV file on the machine running the gradio app.

Example

import gradio as gr

def image_classifier(inp):
    return {'cat': 0.3, 'dog': 0.7}

demo = gr.Interface(
    fn=image_classifier,
    inputs="image",
    outputs="label",
    flagging_callback=gr.SimpleCSVLogger()
)

ClassicCSVLogger

gr.ClassicCSVLogger(simplify_file_data=True)

The classic implementation of the FlaggingCallback abstract class in Gradio versions before 5.0. Each flagged sample (both the input and output data) is logged to a CSV file with headers on the machine running the gradio app.

Parameters

simplify_file_data · bool · default: "True"

If True, the file data will be simplified before being written to the CSV file.

Example

import gradio as gr

def image_classifier(inp):
    return {'cat': 0.3, 'dog': 0.7}

demo = gr.Interface(
    fn=image_classifier,
    inputs="image",
    outputs="label",
    flagging_callback=gr.ClassicCSVLogger()
)

CSVLogger

gr.CSVLogger(
    simplify_file_data=True,
    verbose=True,
    dataset_file_name=None
)

The default implementation of the FlaggingCallback abstract class in gradio>=5.0. Each flagged sample (both the input and output data) is logged to a CSV file with headers on the machine running the gradio app. This implementation is concurrent-safe and creates a new dataset file every time the headers of the CSV change.

Parameters

simplify_file_data · bool · default: "True"

If True, the file data will be simplified before being written to the CSV file. If CSVLogger is being used to cache examples, this is set to False to preserve the original FileData class.

verbose · bool · default: "True"

If True, prints messages to the console about the dataset file creation.

dataset_file_name · str | None · default: "None"

The name of the dataset file to be created (should end in “.csv”). If None, the dataset file will be named “dataset1.csv” or the next available number.

Example

import gradio as gr

def image_classifier(inp):
    return {'cat': 0.3, 'dog': 0.7}

demo = gr.Interface(
    fn=image_classifier,
    inputs="image",
    outputs="label",
    flagging_callback=gr.CSVLogger()
)

ChatCSVLogger

class ChatCSVLogger

Flagging callback for chat conversations. Flagged conversations and like/dislike reactions are logged to a CSV file on the machine running the gradio app.

Methods

setup

setup(flagging_dir)

flagging_dir · str · required

Path to the directory where flagged data is stored.

flag

flag(like_data, messages)

like_data · LikeData · required

The like/dislike data from the event.

messages · list · required

The chat messages.

FlagMethod

class FlagMethod

Helper class that contains the flagging options and calls the flagging method. Also provides visual feedback to the user when flag is clicked.

Constructor

FlagMethod(
    flagging_callback,
    label,
    value,
    visual_feedback=True
)

flagging_callback · FlaggingCallback · required

The flagging callback to use.

label · str · required

The label for the flag button.

value · str | None · required

The value associated with this flag option.

visual_feedback · bool · default: "True"

Whether to provide visual feedback when the flag button is clicked.