---
license: mit
task_categories:
- question-answering
language:
- sa
- en
pretty_name: Ramayana Embeddings
---


**Ramayana Embedding Dataset**  
This repository contains a specialized embedding dataset of the ancient Indian epic, Ramayana, suitable for various NLP tasks, semantic search, and text retrieval purposes.

Dataset relesased by [Mercity AI](https://mercity.ai)!

**Dataset Source**  
The original textual data has been sourced from the repository: [Sanskrit Sahitya Data Repository  
](https://github.com/sanskritsahitya-com/data)

We extend our sincere gratitude to the maintainers of this repository for compiling and sharing valuable Sanskrit literature datasets openly.

**Embedding Models Used**  
The textual embeddings provided in this dataset have been generated using the following embedding models:

 -  [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5)

    Type: Transformer-based sentence embedding model  
    Usage: Particularly suited for semantic search, clustering, and information retrieval tasks.
    
 -  [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2)
   
    Type: Transformer-based sentence embedding model  
    Usage: Efficient and lightweight embeddings suitable for general semantic search and retrieval tasks.
    

**Dataset Structure**  
Each entry in the dataset contains:

 - id: A unique identifier for each verse/excerpt.
 - sanskrit_text: The original textual content in Sanskrit from Ramayana.  
 - translation: English translation of the Sanskrit text.  
 - minilm_embedding: Vector representation (384-dimensional embedding) generated using the [sentence-transformers/all-MiniLM-L6-v2  
    ](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2)
  - bge_embedding: Vector representation (1024-dimensional embedding) generated using the [BAAI/bge-large-en-v1.5  
    ](https://huggingface.co/BAAI/bge-large-en-v1.5)