nassimjp/Persian-Pashto-Cache

Persian-Pashto Translation Cache

This dataset contains parallel translation pairs between Persian (fa) and Pashto (ps), structured for efficient lookup and dataset curation.

Data Structure

The data is stored in JSON Lines (.jsonl) format containing source-target string pairs:

{"persian": "...", "pashto": "..."}

Python Conversion Snippet

The dataset was formatted using the following script:

import json

# Load your dictionary file
with open("translation_cache.json", "r", encoding="utf-8") as f:
    data = json.load(f)

# Convert to list of records (e.g., source / target pairs)
formatted_data = [
    {"persian": src, "pashto": tgt} 
    for src, tgt in data.items()
]

# Save as .jsonl (Recommended for HF Viewer)
with open("data/translations.jsonl", "w", encoding="utf-8") as f:
    for entry in formatted_data:
        f.write(json.dumps(entry, ensure_ascii=False) + "\n")