---
language:
- te
license: apache-2.0
task_categories:
- text-generation
- question-answering
pretty_name: Telugu Instruction Dataset — Luuka AI (10x Technologies)
size_categories:
- 1K<n<10K
tags:
- telugu
- instruction-tuning
- voice-assistant
- luuka
- 10x-technologies
- low-resource-language
- southern-indian-languages
configs:
- config_name: conversational
  data_files:
  - split: train
    path: File-1_Conversational.jsonl
- config_name: emotional_responses
  data_files:
  - split: train
    path: File-2_Emotional Responses.jsonl
- config_name: rag
  data_files:
  - split: train
    path: File-3_RAG.jsonl
- config_name: safety_categories
  data_files:
  - split: train
    path: "File-4 All 600 Categories.jsonl"
- config_name: identity
  data_files:
  - split: train
    path: File-5_200-Identity.jsonl
- config_name: multiturn
  data_files:
  - split: train
    path: File-6_Multi Turn.jsonl
- config_name: helpful
  data_files:
  - split: train
    path: File-7_Helpful.jsonl
- config_name: neutral
  data_files:
  - split: train
    path: File-8_Neutral.jsonl
- config_name: language_quality
  data_files:
  - split: train
    path: "File-9 All LangQual 600.jsonl"
- config_name: time_context
  data_files:
  - split: train
    path: File-10_Time Context.jsonl
dataset_info:
- config_name: conversational
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 1999
- config_name: emotional_responses
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 799
- config_name: rag
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 1494
- config_name: safety_categories
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 599
- config_name: identity
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 199
- config_name: multiturn
  features:
  - name: turns
    sequence:
      struct:
      - name: instruction
        dtype: string
      - name: response
        dtype: string
  num_rows: 399
- config_name: helpful
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 999
- config_name: neutral
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 410
- config_name: language_quality
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 399
- config_name: time_context
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  num_rows: 199
---

# Telugu Instruction Dataset — Luuka AI

**Built by [10x Technologies](https://huggingface.co/10xtechnologieS)** — a curated Telugu-language instruction-tuning dataset for training the **Luuka** voice assistant.

---

## Overview

This dataset contains **7,496 instruction–response pairs** and **399 multi-turn conversations** in Telugu, covering a broad range of natural voice assistant interactions. Every response is written entirely in Telugu script — no English characters appear in any response field.

| Subset | Config Name | Pairs | Description |
|---|---|---|---|
| Conversational | `conversational` | 1,999 | Greetings, small talk, everyday chitchat |
| Emotional Responses | `emotional_responses` | 799 | Empathetic, supportive replies |
| RAG | `rag` | 1,494 | Context + question → grounded answers |
| Safety Categories | `safety_categories` | 599 | Refusals and boundary-setting across 600 harm categories |
| Identity | `identity` | 199 | Luuka self-identification, company info |
| Multi-turn | `multiturn` | 399 | Two-turn conversations teaching context continuity |
| Helpful | `helpful` | 999 | Task assistance, how-to, device control |
| Neutral | `neutral` | 410 | Balanced, factual responses |
| Language Quality | `language_quality` | 399 | High-quality Telugu fluency examples |
| Time Context | `time_context` | 199 | Date, time, and temporal awareness |

**Total: ~6,496 flat pairs + 399 multi-turn conversations**

---

## Dataset Structure

### Flat pairs (all subsets except `multiturn`)

```json
{
  "instruction": "భారత రాజధాని ఏది?",
  "response": "న్యూఢిల్లీ మన భారతదేశ రాజధాని అండీ."
}
```

### Multi-turn (`multiturn` subset)

```json
{
  "turns": [
    {"instruction": "హాయ్", "response": "నమస్తే అండీ! నేను లూకా."},
    {"instruction": "మీరు ఎవరు?", "response": "నేను టెన్ ఎక్స్ టెక్నాలజీస్ తయారు చేసిన లూకాను."}
  ]
}
```

---

## Usage

```python
from datasets import load_dataset

# Load a specific subset
ds = load_dataset("10xtechnologieS/telugu_instruction_dataset", "conversational")
print(ds["train"][0])

# Load multi-turn conversations
ds_mt = load_dataset("10xtechnologieS/telugu_instruction_dataset", "multiturn")
print(ds_mt["train"][0])
```

### Fine-tuning example

```python
from datasets import load_dataset, concatenate_datasets

subsets = [
    "conversational", "emotional_responses", "rag",
    "helpful", "neutral", "language_quality",
    "identity", "time_context"
]

splits = [
    load_dataset("10xtechnologieS/telugu_instruction_dataset", s)["train"]
    for s in subsets
]

combined = concatenate_datasets(splits)
print(f"Total examples: {len(combined)}")
```

---

## Design Principles

- **Telugu-only responses** — English letters (a–z, A–Z) never appear in any response field
- **Voice-first phrasing** — responses are natural for text-to-speech output
- **Safety coverage** — 600 harm categories explicitly handled with appropriate refusals
- **Context continuity** — multi-turn pairs teach the model to carry context across turns
- **Luuka persona** — consistent identity as a 10x Technologies voice assistant throughout

---

## About Luuka

Luuka is a Telugu-language voice assistant developed by **10x Technologies**. This dataset was created to fine-tune Luuka for natural, culturally appropriate, and safe Telugu conversations.

---

## License

[Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)

---

## Citation

```bibtex
@dataset{10xtechnologies2025telugu,
  title        = {Telugu Instruction Dataset — Luuka AI},
  author       = {10x Technologies},
  year         = {2025},
  publisher    = {Hugging Face},
  url          = {https://huggingface.co/datasets/10xtechnologieS/telugu_instruction_dataset}
}
```