vedantjadhav701/edu-qwen-1.7b-merged

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 10, 2026License:mitArchitecture:Transformer Open Weights Featherless Exclusive Cold

Vidya 1.7B, developed by vedantjadhav701, is a 1.7 billion parameter open-source multilingual educational language model. Fine-tuned for 11 Indian languages and focused on NCERT curriculum, it achieves an overall accuracy of 93.3% on the Vidya Multilingual Educational Evaluation Suite. This model excels in STEM domains like Chemistry (99.4%) and Physics (95.6%), making it suitable for educational applications requiring high accuracy in Indian languages.

Loading preview...

Vidya 1.7B: Multilingual Educational LLM

Vidya 1.7B is an open-source, 1.7 billion parameter language model developed by vedantjadhav701, specifically fine-tuned for educational applications. Its primary focus is on the NCERT curriculum across 11 Indian languages, making it a specialized tool for educational content generation and understanding in a multilingual Indian context.

Key Capabilities & Performance

The model's performance was rigorously evaluated using the Vidya Multilingual Educational Evaluation Suite (v1.0), which includes 64 questions across 8 Indian writing systems and 4 STEM domains. Vidya 1.7B achieved an impressive overall accuracy score of 93.3%.

  • Exceptional STEM Performance: Demonstrates near-perfect accuracy in Chemistry (99.4%) and exceptional accuracy in Physics (95.6%) and Biology (95.6%). Mathematics shows good reasoning at 82.5% accuracy.
  • Strong Multilingual Support: Achieves high accuracy across 11 Indian languages, including English (97.5%), Marathi (95.0%), Hindi, Maithili, Tamil, Urdu (92.5%), Telugu, and Bengali (91.9%).

Use Cases

This model is particularly well-suited for:

  • Educational Content Generation: Creating or assisting with content related to the NCERT curriculum in multiple Indian languages.
  • Multilingual Tutoring Systems: Powering AI tutors or educational chatbots that can interact in various Indian languages.
  • Academic Question Answering: Providing accurate answers to questions in STEM subjects and general knowledge, especially within the Indian educational framework.

Finetunes
1 models