BharatGPT group led by IIT Bombay to launch ‘Hanooman’ a ChatGPT-like service | Current Affairs | Vision IAS
MENU
Home

Periodically curated articles and updates on national and international developments relevant for UPSC Civil Services Examination.

Quick Links

High-quality MCQs and Mains Answer Writing to sharpen skills and reinforce learning every day.

Watch explainer and thematic concept-building videos under initiatives like Deep Dive, Master Classes, etc., on important UPSC topics.

ESC
  • Hanooman is a series of Large Language Models (LLMs) that can respond in various Indian languages like Hindi, Tamil, and Marathi.
    • It is expected to work in four fields, including health care, governance, financial services, and education.
  • LLMs are deep learning models that consume and train on massive datasets to excel in language processing tasks
    • They create new combinations of text that mimic natural language based on its training data.
    •  The training data is usually large datasets including Wikipedia and OpenWebText. 

 

  • Applications of LLM
    • Medical:  For tasks like protein structure prediction that can uncover patterns in disease and predict outcomes.
    • Retail: To provide enhanced customer experiences through dynamic chatbots. 
    • Software: To write software and teach robots how to do physical tasks.
    • Finance: To summarize earnings calls and create transcripts of important meetings. 
    • Marketing: To organize customer feedback. 

 

  • Challenges in developing LLM: Significant capital investment, large datasets, technical expertise, and large-scale compute infrastructure necessary to develop and maintain LLMs. 
    • Sourcing of quality datasets in Indian languages is a major hurdle in developing LLM in Indian languages.

Generative Pretrained Transformer (GPT) vs LLM

  • A GPT is a type of LLM that uses deep learning to generate human-like text. 
  • They are called 
    • "generative" because they can generate new text based on the input they receive, 
    • "pretrained" because they are trained on a large corpus of text data before being fine-tuned for specific tasks, and 
    • "transformers" because they use a transformer based neural network architecture to process input text and generate output text.
Watch Video News Today

Explore Related Content

Discover more articles, videos, and terms related to this topic

Title is required. Maximum 500 characters.

Search Notes

Filter Notes

Loading your notes...
Searching your notes...
Loading more notes...
You've reached the end of your notes

No notes yet

Create your first note to get started.

No notes found

Try adjusting your search criteria or clear the search.

Saving...
Saved

Please select a subject.

Referenced Articles

linked

No references added yet

Subscribe for Premium Features