Downloads · 30 days
34
8% of all-time downloads
deberta-v3-base-japanese-politeness
namopanda/deberta-v3-base-japanese-politeness
deberta-v3-base-japanese-politeness is a text classification model from namopanda. Use it when you need a label for a piece of text. The card lists the license as mit.
This model was fine-tuned on microsoft/deberta-v3-base using the KeiCO-corpus (https://github.com/Liumx2020/KeiCO-corpus). KeiCO-corpus © Ochanomizu University Kobayashi Lab, Liu Muxuan, licensed under CC BY 4.0 (http…
All-time downloads
404
Public
Parameters
184M
1.5 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin738 MB · 50%
Parameter types
How the weights are stored.
F32184M · 100%
Base models
Model card
From the Hugging Face model README
Model Overview
This model was fine-tuned on microsoft/deberta-v3-base using the KeiCO-corpus (https://github.com/Liumx2020/KeiCO-corpus). KeiCO-corpus © Ochanomizu University Kobayashi Lab, Liu Muxuan, licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).
We trained the model on the dataset split 8:1:1, achieving an accuracy of 0.747.
There are four levels, with politeness increasing as the level decreases, and LABEL_0 being the most polite.
Usage
from transformers import pipeline
classifier = pipeline("text-classification", model="namopanda/deberta-v3-base-japanese-politeness")
text = "こんにちは。お元気ですか？"
output = classifier(text)
print(f"label: {output[0]["label"]} score: {output[0]["score"]}")
- Task
- Text Classification
- Type
- deberta-v2
- License
- mit
- Languages
- ja
- Created
- Nov 14, 2025
- Updated
- Dec 14, 2025