nassimjp/zamai-pashto-clean-cpt
ZamAI Pashto Clean CPT This dataset is a hyper-cleaned, optimized, and fully deduplicated version of tasal9/ZamAI-Pashto-Mega-Dataset intended for Causal Language Modeling (CLM), pre-training, or fine-tuning text models in the Pashto language. 🛠️ Pipeline & Filtering Details Before processing the ~1.5 GB stream, an Internal Built-In Self-Test (BIST) was executed to verify environment I/O permissions, validate character ratio logic, and check connection… See the full description on the dataset page: https://huggingface.co/datasets/nassimjp/zamai-pashto-clean-cpt.
ZamAI Pashto Clean CPT
This dataset is a hyper-cleaned, optimized, and fully deduplicated version of tasal9/ZamAI-Pashto-Mega-Dataset intended for Causal Language Modeling (CLM), pre-training, or fine-tuning text models in the Pashto language.
🛠️ Pipeline & Filtering Details
Before processing the ~1.5 GB stream, an Internal Built-In Self-Test (BIST) was executed to verify environment I/O permissions, validate character ratio logic, and check connection constraints.
The processing pipeline applied the following steps:
- URL & Email Stripping: Removed web artifacts, URLs (
http/https/www), and electronic mail patterns. - Whitespace Normalization: Truncated consecutive tabs, newlines, and spaces into single tokens.
- Character Ratio Validation: Calculated the exact density of native Pashto scripts—safeguarding unique characters like
ښandځ. Items below a0.2ratio threshold were systematically filtered out. - MD5 Deduplication: Maintained a flat execution hash set to drop completely identical passages.
📊 Dataset Statistics
🚀 Usage
You can load this optimized dataset seamlessly into your PyTorch/Hugging Face training pipelines:
from datasets import load_dataset
dataset = load_dataset("nassimjp/zamai-pashto-clean-cpt", split="train")
# Preview the first sample
print(dataset[0])