Datasetpublic

rodriguescarson/adaption-konkani-script-12k

Konkani Script Conversion Konkani script tasks: Devanagari to Kannada script, Kannada script back to Devanagari, ISO 15919 romanisation, and script identification. Rows 12,000 Domain Konkani language Format data.parquet, one row per example Licence cc-by-4.0 Built for supervised fine-tuning (SFT) experiments on Adaption AutoScientist Columns Column Description original_prompt The prompt (user turn) as uploaded.… See the full description on the dataset page: https://huggingface.co/datasets/rodriguescarson/adaption-konkani-script-12k.

sourceHugging Facecc-by-4.0updated 14d agoView on Hugging Face

0likes68downloads

Dataset Card

Konkani Script Conversion

Konkani script tasks: Devanagari to Kannada script, Kannada script back to Devanagari, ISO 15919 romanisation, and script identification.


Rows	12,000
Domain	Konkani language
Format	`data.parquet`, one row per example
Licence	cc-by-4.0
Built for	supervised fine-tuning (SFT) experiments on Adaption AutoScientist

Columns

Column	Description
`original_prompt`	The prompt (user turn) as uploaded.
`original_completion`	The target response as uploaded.
`enhanced_prompt`	Empty in this dataset.
`enhanced_completion`	Empty in this dataset.

How it was built

6,025 verified Konkani sentences from ai4bharat/sangraha, each kept only if its script conversions round-trip exactly; 9,000 conversion labels re-derived after writing with 0 errors.

Sources and licence

—ai4bharat/sangraha (CC-BY-4.0)

Notes

—Columns enhanced_prompt, enhanced_completion are empty in this dataset (the Adaption export reserves them for rewritten text).

Loading

python

from datasets import load_dataset
ds = load_dataset("rodriguescarson/adaption-konkani-script-12k", split="train")

python

import pandas as pd
df = pd.read_parquet("hf://datasets/rodriguescarson/adaption-konkani-script-12k/data.parquet")

Published by Carson Rodrigues (Hugging Face, Kaggle).