---
license: apache-2.0
language:
- hi
metrics:
- perplexity
widget:
- text: "BCCI ने टी-20 वर्ल्ड कप के बीच जिम्बाब्वे सीरीज के लिए टीम इंडिया का ऐलान कर दिया है। इस टीम में कई नए चेहरों को जगह दी गई है।"
  example_title: "Example 1"
- text: "7 अक्टूबर को हमास से जंग शुरू होने के सात महीने बाद इजरायली सेना ने गाजा पट्टी में हमास के ठिकानों पर हमला किया था। इस हमले में हमास के कई ठिकानों को निशाना"
  example_title: "Example 2"
---

# Model Card for Ganga-1b! 🌊

The base model **``Ganga-1b``** trained on a monolingual **Hindi** language dataset as part of ***Project Unity***. The first pre-trained Hindi model by any academic research lab in India 🇮🇳! 


![image/png](https://cdn-uploads.huggingface.co/production/uploads/667b8f8ba271fc5a8e6929de/jG3tZnGPvH6vcGrvxO-YC.png)


## Model Details

### Model Description

Project Unity is an initiative aimed at addressing India's linguistic diversity and richness by creating a comprehensive resource that covers the country's major languages. Our goal is to achieve state-of-the-art performance in understanding and generating text in Indian languages. To achieve this, we train models on monolingual regional languages of India. Our first release is the Ganga-1B model, which has been trained on a large dataset of public domain web-crawled hindi language data, including news articles, web documents, books, government publications, educational materials, and social media conversations (filtered for quality). Additionally, the dataset has been further curated by native Indian speakers to ensure high-quality. Importantly, the Ganga-1B model outperforms existing open-source models that support Indian languages, even at sizes of up to 7 billion parameters. Designed to be compact and efficient, the model can easily run on edge devices, making it ideal for a range of applications that require generating human-like text. Its modest size also enables easy integration into resource-constrained environments, such as personal devices or cloud infrastructure, allowing for wider adoption and innovation in AI-driven technologies.



- **Developed by:** Lingo Research Group and hyperlink 
- **Model type:** Transformer-based Language Model
- **Language(s) (NLP):** Bilingual (Primary: Hindi [hi], Secondary: English [en]
- **License:** Apache 2.0



## How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]


## Bias, Risks, and Limitations


### Recommendations

This model described is a research preview and is under ongoing iterative updations, and as such, it only provides limited safety measures. Additionally, it may generate offensive content. It is strictly prohibited to use this service for any illegal, harmful, violent, racist, or sexual purposes.





## Evaluation



### Results


|    Model    | Fertility |   PPL  |
|:-----------:|:---------:|:------:|
|   ganga-1b  |    1.12   |  34.85 |
|  pragna-1b  |    1.58   |  12.74 |
|  bloom-1b1  |    1.27   |  33.39 |
|  bloom-1b7  |    1.27   |  26.63 |
| airavata-7b |    1.69   |  46.24 |
|   bloom-3b  |    1.27   |  23.77 |
|   gemma-2b  |    1.89   |  41.67 |


#### Summary


## Technical Specifications

### Model Architecture and Objective


Ganga-1b is a decoder-only transformer model, featuring the following specifications:


* #Layers: 16
* #Attention heads: 32
* Embedding dimension: 2048
* Vocabulary size: 30000
* Sliding window: 512
* Intermediate dimension: 716


## Model Card Contact

Lingo-IITGN