Datasets:

Azri-Muhsin
/

sangraha-tamil-cleaned

language:
  - ta
license: cc-by-4.0
task_categories:
  - text-generation
  - language-modeling
tags:
  - nlp
  - tamil
  - sangraha
  - continuous-pretraining
size_categories:
  - 1M<n<10M

Sangraha Tamil (Cleaned & Verified)

Dataset Description

This dataset is a cleaned, processed subset of the AI4Bharat Sangraha (Verified) dataset, specifically targeting the Tamil language. It was prepared for the purpose of Continuous Pre-Training (CPT) of Large Language Models (LLMs) like Llama 3 and Qwen 2.5/3 to improve their performance on Indic languages.

Original Source: AI4Bharat Sangraha (Verified Split)
Language: Tamil (ta)
Format: Parquet
Approximate Size: ~7.8 Million Rows (Processed)

Data Processing Pipeline

To ensure high-quality training tokens, the following cleaning heuristics were applied to the raw data: (most heavyduty preprocessing tasks were handled by the AI4Bharat/Sangraha folks )

Unicode Normalization: Applied NFC normalization.
HTML Removal: All HTML tags and artifacts stripped.
Script Density Filter: Rows were dropped if Tamil characters constituted < 50% of the text. This filters out English-heavy content mixed into the dataset.
Length Filtering: Sentences with fewer than 20 characters were removed to reduce noise.
Heuristic Noise Filtering: Removed common web-scrape noise phrases (e.g., "home screen", "click here", "amazon card").
Deduplication: Relied on the original Sangraha "Verified" pipeline (MinHash LSH).

Usage

Since the dataset is large, it is recommended to use streaming=True or load it via load_dataset in chunks.

from datasets import load_dataset

# Load in streaming mode (Recommended for local PCs)
dataset = load_dataset("Azri-Muhsin/sangraha-tamil-cleaned", split="train", streaming=True)

for sample in dataset.take(3):
    print(sample["text"])