BhashaBench-V1 is a domain-specific, multi-task, multilingual benchmark designed to evaluate large language models in India-centric contexts.
AI & ML interests
None defined yet.
Recent Activity
Papers
IndicParam: Benchmark to evaluate LLMs on low-resource Indic Languages
AyurParam: A State-of-the-Art Bilingual Language Model for Ayurveda
models 15
bharatgenai/Param-1-2.9B-Instruct
Text Generation • 3B • Updated
• 1.68k • 17
bharatgenai/Shrutam-2
Automatic Speech Recognition • Updated
bharatgenai/AyurParam
Text Generation • 3B • Updated
• 2.29k • 17
bharatgenai/FinanceParam
Text Generation • 3B • Updated
• 170 • 5
bharatgenai/LegalParam
Text Generation • 3B • Updated
• 537 • 7
bharatgenai/AgriParam
Text Generation • 3B • Updated
• 341 • 7
bharatgenai/Param-1
Text Generation • Updated
• 201 • 6
bharatgenai/sooktam2
Text-to-Speech • Updated
• 1
bharatgenai/Param2-17B-A2.4B-Thinking
Text Generation • 17B • Updated
• 570 • 20
bharatgenai/Param-1-7B
7B • Updated
• 39 • 2
datasets 7
bharatgenai/IndicParam
Viewer
• Updated
• 13.2k • 90 • 6
bharatgenai/ParamBench
Viewer
• Updated
• 17.3k • 24 • 4
bharatgenai/TORQUE
Viewer
• Updated
• 210 • 206 • 1
bharatgenai/BhashaBench-Legal
Viewer
• Updated
• 24.4k • 158 • 4
bharatgenai/BhashaBench-Finance
Viewer
• Updated
• 19.4k • 187 • 2
bharatgenai/BhashaBench-Krishi
Viewer
• Updated
• 15.4k • 68 • 6
bharatgenai/BhashaBench-Ayur
Viewer
• Updated
• 15k • 89 • 9