mradermacher/Math-AI-Full-GGUF
4B โข Updated โข 148 โข 1
Datasets:
1
๐ ShORT-สษชษดษขสษชsส-๐ฏ๐๐๐๐๐๐๐-๐ท๐ถแด ๐ฎ๐ณ
SKT AI LABS
The Sovereign LLM Development For India (Project Surya)
โจ Overview
This dataset is a monumental collection of 10 Million high-quality conversation pairs crafted in Hinglish (Hindi + English). It is meticulously engineered to empower Large Language Models (LLMs) with a deep understanding of Indian linguistic nuances and conversational contexts.
๐ ๏ธ Dataset Details
- ๐จโ๐ป Developer: Shrijan Kumar Tiwari & SKT AI Labs Research Team
- ๐ Project Name:
ShORT-Hinglish-Dataset-10M - ๐ Format:
JSONL(JSON Lines) - ๐ข Total Records: 10,000,000 (10 Million)
- ๐ Topics Covered: Coding, Science, History, Mathematics, Geography, Space and Technology.
๐ Key Features
- โ High Accuracy: Every question is strictly mapped to a verified and factually correct response.
- ๐ Diversity: Built with multiple response templates to prevent repetitive patterns and mode collapse.
- ๐ Clean Data: Each record features a unique UUID and rich metadata for seamless tracking and filtering.
- ๐ฎ๐ณ Localization: Perfect balance of Hindi and English for native-feel AI interactions.
๐ Use Cases & Usage
- ๐ค Model Training: Ideal for fine-tuning or building new LLMs from scratch.
- ๐ Education & Tech: Perfect for training specialized chatbots in academic and technical domains.
- ๐ Linguistic Research: Great for studying code-switching patterns in South Asian languages.
๐ Licensing & Authorship
- โ๏ธ License: Apache 2.0
- โ๏ธ Authors: Created by the SKT AI TEAM under the leadership of Shrijan Kumar Tiwari.
Note: For future conversations, Contact US --
- Downloads last month
- 24
