LLM Evaluation · தமிழ்
Tamil Data for LLM Evaluation
Human evaluation of large language model output in Indian languages, including cultural and factual fit. In Tamil, the binding constraint is usually dialect coverage and code-mixing, not raw hours.
- Language
- Tamil
- Primary metric
- Rubric scores with confidence intervals
- Typical volume
- 500-2,000 hours
Data profile required
- Native-speaker rater panels per language
- Rubric-based scoring with calibration
- Overlapping assignments for agreement
What Tamil adds to the requirement
- Extreme diglossia: written Tamil and spoken Tamil differ so much that read-speech corpora are near-useless for conversational ASR
- Tamil script under-specifies voicing, so க can surface as /k/, /g/, /h/ or /x/ depending on position
- Dialects to cover: Chennai (Madras Bashai), Kongu (Coimbatore), Madurai, Nellai (Tirunelveli)
- Tanglish is the default urban register. Technology, finance, and workplace vocabulary is largely English embedded in Tamil syntax.
Metrics to track
- Rubric scores with confidence intervals
- Inter-rater agreement
- Failure-mode distribution
Failure modes
- Raters who are fluent but not native in the variety
- Rubrics written in English and applied to non-English output without localisation
For Tamil specifically: Almost all public Tamil audio is literary read speech from news or scripture. Genuine colloquial Tamil, especially southern and Kongu varieties, is the single biggest gap for anyone building Tamil voice products.
Recommended cohort
Recruit by district rather than by city alone; Chennai-only cohorts produce models that degrade sharply in the south and west of the state.
| Dimension | Typical split | Why it matters for Tamil |
|---|---|---|
| Gender | 50 / 50 | Pitch range differences change acoustic model behaviour; unbalanced cohorts bias recognition |
| Age | 18-25: 30%, 26-40: 40%, 41-60: 30% | Older speakers retain conservative Tamil forms that younger urban speakers have lost |
| Region | Tamil Nadu / Puducherry / parts of Karnataka and Kerala and others | Dialect spread across 6 recognised varieties |
| Education | Mixed, including below-graduate | Prompt-reading fluency correlates with education and skews prosody |
| Condition | Studio / quiet room / field | Match the noise profile of your deployment |
Suggested programme shape
Start with an evaluation set of 120 speakers spread across every Tamil dialect in scope, collected before training data. Then field 500-2,000 hours of training data from disjoint speakers.
This ordering is what makes the improvement measurable rather than assumed.
Frequently asked
Is there usable public Tamil data for llm evaluation?
Almost all public Tamil audio is literary read speech from news or scripture. Genuine colloquial Tamil, especially southern and Kongu varieties, is the single biggest gap for anyone building Tamil voice products.
How many Tamil speakers do we need?
1,000-3,000 speakers for a training corpus, plus a disjoint evaluation cohort covering each dialect. Speaker count matters more than hours for generalisation.
Can you run this across multiple languages at once?
Yes. Multi-language programmes run to one master specification so per-language results stay comparable.
Scope Tamil data for llm evaluation
Send the target metric and the languages in scope.