Downloads · 30 days
9
3% of all-time downloads
sanskrita-gpt2
harshalJk/sanskrita-gpt2
sanskrita-gpt2 is a machine learning model from harshalJk. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Sanskrita is a GPT-2 based language model fine-tuned on IAST-transliterated Sanskrit texts and enhanced with a Retrieval-Augmented Generation (RAG) pipeline for scripture-grounded question answering.
All-time downloads
335
Public
Parameters
124M
996 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors498 MB · 100%
At a glance
- Model type
- gpt2
- Access
- Public
- Created
- Feb 6, 2026
- Updated
- Feb 7, 2026
- SHA
- ad7c3ae3
Model card
From the Hugging Face model README
Sanskrita: Retrieval-Augmented GPT-2 for Sanskrit Q&A
Overview
Sanskrita is a GPT-2 based language model fine-tuned on IAST-transliterated Sanskrit texts and enhanced with a Retrieval-Augmented Generation (RAG) pipeline for scripture-grounded question answering.
Training Data
- GURU Dataset (Digital Corpus of Sanskrit – DCS)
- Krishna–Arjuna Q&A dataset (Bhagavad Gita inspired dialogue)
Model Details
- Base model: GPT-2
- Tokenization: Custom tokenizer for IAST Sanskrit
- Max context length: 1024
Intended Use
- Sanskrit question answering
- Philosophical dialogue generation
- Academic & educational research
Limitations
- Not trained on Devanagari
- Faithfulness depends on retrieval quality
Citation
- Type
- gpt2
- Created
- Feb 6, 2026
- Updated
- Feb 7, 2026