📘 Overview

The DebertaV3SampleEvaluator is a text quality classifier based on the Nvidia Deberta V3 model. It is used to evaluate the quality of a given text and return a classification result.

__init__

def __init__(self, model_name, model_cache_dir='./dataflow_cache', device='cuda')

Parameter	Type	Default	Description
model_name	str	Required	The name of the pretrained model to use.
model_cache_dir	str	'./dataflow_cache'	The directory where the model is cached.
device	str	'cuda'	The device to run the model on (e.g., 'cuda', 'cpu').

Prompt Template Descriptions

run

def run(self, storage: DataFlowStorage, input_key: str, output_key: str='Debertav3Score')

Parameter	Type	Default	Description
storage	DataFlowStorage	Required	The DataFlowStorage instance for reading and writing data.
input_key	str	Required	The name of the input column containing the text to be evaluated.
output_key	str	'Debertav3Score'	The name of the output column to store the classification result.

🧠 Example Usage

🧾 Default Output Format

Field	Type	Description
...	any	Original columns from the input DataFrame.
Debertav3Score	str	The quality classification result from the model. The actual column name is determined by the output_key.