---
annotations_creators: []
language:
  - en
language_creators: []
license: apache-2.0
multilinguality:
  - monolingual
pretty_name: adaption-legalbrain_indic_legal
size_categories:
  - 10K<n<100K
source_datasets:
  - extended|https://huggingface.co/datasets/Prarabdha/indian-legal-supervised-fine-tuning-data

tags:
  - adaption
  - legal
  - indian-law
  - legal-ai
  - instruction-tuning
  - llm
  - question-answering

task_categories:
  - question-answering
  - text-generation

task_ids:
  - open-domain-qa
  - text2text-generation

---

![banner](https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/bf1fda4c-e045-4344-90ef-9c752985e599.png)

# Adaption LegalBrain Indic Legal

**Adaption LegalBrain Indic Legal** is a professionally remastered version of the original **Indian Legal Supervised Fine-Tuning Dataset**. The dataset has been enhanced using **Adaption's Adaptive Data Platform**, improving instruction quality, consistency, and training effectiveness for Legal AI applications.

**Original Dataset:** https://huggingface.co/datasets/Prarabdha/indian-legal-supervised-fine-tuning-data

---

# Overview

This dataset contains high-quality instruction-response pairs covering Indian legal concepts, constitutional law, criminal law, contract law, corporate law, procedural law, and related judicial topics.

It is designed for:

- Legal Question Answering
- Supervised Fine-Tuning (SFT)
- Legal Chatbots
- AI Legal Assistants
- Retrieval-Augmented Generation (RAG)
- Legal Reasoning Systems
- Instruction-Tuned Large Language Models

The dataset has been automatically optimized using Adaption's Adaptive Data pipeline, resulting in improved prompt quality, reduced redundancy, and better instructional consistency.

---

# Dataset Statistics

| Property | Value |
|----------|------:|
| Dataset Type | Instruction Tuning |
| Total Samples | **32,105** |
| Language | English |
| Domain | Indian Legal AI |
| Quality Grade | **B** |
| Relative Quality Improvement | **78.0%** |

---

# Domain Distribution

- Legal — **92%**
- Corporate & Business — **4%**
- Religion — **2%**

---

# Writing Style

- Informative — **44%**
- Professional — **20%**
- Analytical — **20%**

---

# Adaptive Data Evaluation

### Quality Improvement

<img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/b0a9fba5-5c2d-4cf8-8e88-cc638c2106fc.png" width="650"/>

---

### Grade Improvement

<img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/06f10dfe-6178-47f4-875d-a348456fcdb4.png" width="650"/>

---

### Percentile Improvement

<img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/1a3ec266-5825-4fa2-908c-eb7e60eb3d23.png" width="650"/>

---

# Key Features

- High-quality legal instruction-response pairs
- Optimized using Adaption Adaptive Data
- Suitable for supervised fine-tuning
- Consistent formatting across samples
- Reduced prompt duplication
- Enhanced instruction clarity
- Ready for Legal LLM training
- Ideal for legal research assistants and AI chatbots

---

# Recommended Use Cases

- Fine-tuning open-source LLMs
- Legal AI Assistants
- Constitutional Law QA
- Court Judgment Analysis
- Legal Education
- Legal Search Systems
- Retrieval-Augmented Generation (RAG)
- Academic Research

---


# Credits

**Dataset Publisher**

**Nupur Saraswat**

Mastered using **Adaption Adaptive Data Platform**

---

# Citation

```bibtex
@dataset{adaption_legalbrain_indic_legal,
  title={Adaption LegalBrain Indic Legal},
  author={Nupur Saraswat},
  year={2026},
  publisher={Hugging Face},
  note={Remastered using Adaption Adaptive Data Platform. Original dataset by Prarabdha.}
}
```