Family
IndicBERTv2-LoRA + ONNX int8
productiontaintedtextretail-facing
Sentiment Analyzer.
4-class financial-chat sentiment classifier (distressed/negative/positive/neutral) serving /ai/ask via inline CPU ONNX int8.
Foundation · IndicBERTv2-LoRA + ONNX int8
Marketing claim
Every foundation model in the inference path is Apache 2.0 or MIT — no Llama, no Claude, no GPT, no Gemini.Note ·All foundation weights Apache 2.0 or MIT; one distilled adapter (sentiment_analyzer) is undergoing replacement — see /models/sentiment_analyzer.
What it does
The honest description.
4-class financial-chat sentiment classifier (distressed/negative/positive/neutral) serving /ai/ask via inline CPU ONNX int8.
Architecture
How it works.
Free-text user message in any of 8 supported languages (EN/HI/TA/TE/KN/BN/MR/UR). 4-way sentiment label (distressed / negative / positive / neutral) + confidence [0,1] + source (sentiment_analyzer / rule_based on fallback).
Training data & drift
What it learned from, and how we watch it.
Source
Inputs: Free-text user message in any of 8 supported languages (EN/HI/TA/TE/KN/BN/MR/UR).
Drift baseline · current quarter
Drift baseline not yet documented for this model.
What it's NOT
- Legacy distilled adapter undergoing planned replacement under our foundation-model license policy. The replacement is a clean Apache-2.0 redistillation; this card is updated when it clears the quality gate.
- This adapter is being replaced; the successor uses only Apache-2.0 / MIT-licensed foundations and must pass our quality gate before promotion.
Who calls this model
Across the house and the API.
Internal routes
No public routes documented yet.
Version history · immutable once published
Every release, traceable.
Data Mukodi holds about you is subject to DPDP §12 erasure (use /consent in the app), and any concern about this model can be raised via [email protected] under DPDP §13 grievance, or via /security.