---
license: apache-2.0
language:
- en
pretty_name: LifeTextMultiTurnStreamingCoT
tags:
- streaming-cot
- life-scenarios
- text
- multi-turn
- dialogue
- sft
- reasoning
- instruction-tuning
task_categories:
- text-generation
- question-answering
task_ids:
- language-modeling
size_categories: 1K<n<10K
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train.parquet
  - split: eval
    path: data/eval.parquet
- config_name: high_quality
  data_files:
  - split: train
    path: data/high_quality_train.parquet
  - split: eval
    path: data/high_quality_eval.parquet
---

# LifeTextMultiTurnStreamingCoT v0.4

**Final professional public release** — clean SFT schema, target field, canonical taxonomy.

## Overview

- **Modality**: Text
- **Turn Type**: Multi Turn
- **Version**: v0.4
- **License**: apache-2.0
- **Language**: English
- **HF Repo**: `skyzhou06/LifeTextMultiTurnStreamingCoT`

## Row Counts

| Split | Rows |
|-------|------|
| Train | 7,955 |
| Eval  | 2,045 |
| High Quality Train | 4,690 |
| High Quality Eval  | 1,233 |
| **Total** | **10,000** |

## Schema (v0.4)

### Top-Level Fields

| Field | Type | Description |
|-------|------|-------------|
| `id` | string | Stable example ID |
| `split` | string | `train` or `eval` |
| `modality` | string | `text` |
| `turn_type` | string | `multi_turn` |
| `input` | object | Input data |
| `streaming` | object | Checkpoints with natural-language reasoning |
| `target` | object | Training target: reasoning, answer, response |
| `taxonomy` | object | Canonical content classification |
| `quality` | object | Quality assessment |
| `source` | object | Provenance |
| `metadata` | object | Release metadata |

### `target`

```json
{
  "reasoning": "Natural-language reasoning summary.",
  "answer": "The final answer.",
  "response": "Reasoning: ...\n\nAnswer: ..."
}
```

- `target.answer` — for answer-only SFT
- `target.response` — for reasoning-augmented SFT

### `taxonomy`

Uses canonical enum categories: `daily_life`, `travel`, `education`, `work_productivity`, `finance_consumer`, `health_wellness_safe`, `tech_support`, `information_extraction`, `creative_planning`, `social_communication`.

### `quality`

```json
{
  "is_high_quality": true,
  "sft_ready": true,
  "natural_reasoning": true,
  "reasoning_quality": "high",
  "taxonomy_quality": "high"
}
```

## Changes from v0.3

- `output` renamed to `target` with `reasoning`, `answer`, `response`
- `streaming.trace` removed from active rows
- `metadata.legacy` blobs removed
- Taxonomy mapped to canonical enum with fixed category/subcategory pairs
- `high_quality` split into `high_quality_train` / `high_quality_eval`
- `metadata.release_version` = `"v0.4"`

## SFT Usage

```python
from datasets import load_dataset

# Default config
ds = load_dataset("skyzhou06/LifeTextMultiTurnStreamingCoT")
# ds["train"]["target"]["response"]  — reasoning + answer
# ds["train"]["target"]["answer"]    — answer only

# High quality config
ds_hq = load_dataset("skyzhou06/LifeTextMultiTurnStreamingCoT", "high_quality")
```

## Source Licenses

Dataset-level license: **apache-2.0**. Individual rows include `source.license` and `source.dataset` fields with source-specific license information.

## Limitations

- Reasoning is rule-based (content-grounded)
- Some answers are brief closing phrases (check `quality.sft_ready`)
- Non-English examples not included

## Citation

If you use this dataset, please cite the original source datasets and this release.