Dataset for Fine-Tuning Code Generation Models: Kannada-English Algorithmic Statements and Python Code
<p dir="ltr">This dataset was designed to fine-tune code-generation models for converting algorithmic statements written in Kannada or English into Python code. It consists of 1,250 triplets of Kannada algorithmic statements, their English translations, and corresponding Python code....
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , , , , , , , , |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| الملخص: | <p dir="ltr">This dataset was designed to fine-tune code-generation models for converting algorithmic statements written in Kannada or English into Python code. It consists of 1,250 triplets of Kannada algorithmic statements, their English translations, and corresponding Python code. The data is stored in JSON format under the following labels:</p><ul><li><b>"kannada text"</b> – Kannada algorithmic statement</li><li><b>"text"</b> – English algorithmic statement</li><li><b>"code"</b> – Expected Python output</li></ul><h4><b>Dataset Highlights: -</b></h4><ul><li><b>Triplet Structure:</b> Kannada algorithmic statements, their English translations, and Python code.</li><li><b>Diverse Algorithmic Concepts:</b> Covers variable assignments, print statements, conditionals, loops, and comparison operations.</li><li><b>Translation Methodology:</b> English statements were generated from Kannada using the Google Translate API (via Python’s deep_translator module).</li></ul><p dir="ltr">This dataset supports Kannada-to-Python and English-to-Python code generation tasks, enabling step-by-step code generation from algorithmic statements.</p> |
|---|