AmanPriyanshu/reasoning-sft-JSON-structuring-and-correcting

---
license: apache-2.0
task_categories:
- text-generation
- question-answering
language:
- en
- ja
tags:
- reasoning
- sft
- chain-of-thought
- tool-calling
- structured-output
- json
- json-repair
- agent
size_categories:
- 100K<n<1M
---

# JSON Structuring and Correcting (Reasoning SFT)

Combined dataset of 508K rows for training LLMs on structured output tasks with reasoning traces, sourced from two datasets:

## Sources

### tool_calling.parquet (488,461 rows)
Converted from [vericava/sft-tool-calling-structured-output-v1](https://huggingface.co/datasets/vericava/sft-tool-calling-structured-output-v1). Multi-turn tool calling and structured output tasks including tool invocations, tool results, and final assistant responses. Includes English and Japanese content.

### json_repair.parquet (19,645 rows)
Converted from [kth8/json-repair](https://huggingface.co/datasets/kth8/json-repair). JSON formatting and repair tasks where the model corrects malformed JSON into valid, properly indented output.

## Format

Each row has three columns:

- **`input`** — list of dicts with role/content conversation turns (system prompt includes tools/schema where applicable)
- **`response`** — response string with `<think>` reasoning block followed by the structured output
- **`source`** — `N/A`

## Conversion

- Tool-calling: system prompt combines model identity + available tools + output schema; tool calls and results preserved as context turns; 1,322 rows dropped (empty targets)
- JSON repair: system + user turns as input, assistant response as output; 0 rows dropped
- Both: short task-focused reasoning injected in think blocks; validated exactly 1 open and 1 close think tag per response

## Usage

```py
from huggingface_hub import hf_hub_download
import pyarrow.parquet as pq
import random

repo = "AmanPriyanshu/reasoning-sft-JSON-structuring-and-correcting"

for name in ["tool_calling.parquet", "json_repair.parquet"]:
path = hf_hub_download(repo_id=repo, filename=name, repo_type="dataset")
table = pq.read_table(path)
print(f"{'='*80}")
print(f"{name}: {len(table)} rows")
print(f"{'='*80}")
i = random.randint(0, len(table) - 1)
row = {col: table.column(col)[i].as_py() for col in table.schema.names}
print(f"\n[source] {row['source']}")
print(f"\n[input] ({len(row['input'])} turns)")
for t in row["input"]:
preview = t["content"][:250] + ("..." if len(t["content"]) > 250 else "")
print(f" {t['role']}: {preview}")
rp = row["response"][:800]
if len(row["response"]) > 800:
rp += "..."
print(f"\n[response]\n{rp}\n")
```

## License

Apache 2.0 — inherited from both source datasets.

许可证：Apache-2.0
任务类别：
- 文本生成
- 问答
语言：
- 英语
- 日语
标签：
- 推理
- 监督微调（SFT）
- 思维链（Chain-of-Thought）
- 工具调用（Tool Calling）
- 结构化输出（Structured Output）
- JSON
- JSON修复
- AI智能体（AI Agent）
规模类别：10万<样本数<100万

# JSON结构化与修复（推理监督微调）

本数据集包含50.8万条数据，用于训练支持推理轨迹的结构化输出任务的大语言模型（LLM），数据源自两个公开数据集：

## 数据源

### tool_calling.parquet（488,461条数据）
源自[vericava/sft-tool-calling-structured-output-v1](https://huggingface.co/datasets/vericava/sft-tool-calling-structured-output-v1)。涵盖多轮工具调用与结构化输出任务，包含工具调用、工具返回结果与最终助手回复，支持英语与日语内容。

### json_repair.parquet（19,645条数据）
源自[kth8/json-repair](https://huggingface.co/datasets/kth8/json-repair)。包含JSON格式修复任务，要求模型将格式错误的JSON修正为合法且缩进规范的输出内容。

## 数据格式

每条数据包含三列：
- **`input`**：包含角色/内容对话轮次的字典列表（系统提示中会按需包含工具与输出schema）
- **`response`**：回复字符串，以`<think>`推理块开头，后接结构化输出内容
- **`source`**：`N/A`（未指定具体来源）

## 数据处理流程

- 工具调用任务：系统提示由模型身份、可用工具与输出schema组合而成；工具调用与结果作为上下文轮次保留；移除1322条无效目标（空目标）数据
- JSON修复任务：以系统与用户轮次作为输入，助手回复作为输出；未移除任何数据
- 两类任务均：在`<think>`块中注入针对任务的简短推理内容；经校验，每条回复恰好包含1个`<think>`起始标签与1个`<think>`闭合标签

## 使用示例

py
from huggingface_hub import hf_hub_download
import pyarrow.parquet as pq
import random

repo = "AmanPriyanshu/reasoning-sft-JSON-structuring-and-correcting"

for name in ["tool_calling.parquet", "json_repair.parquet"]:
path = hf_hub_download(repo_id=repo, filename=name, repo_type="dataset")
table = pq.read_table(path)
print(f"{'='*80}")
print(f"{name}: {len(table)} 条数据")
print(f"{'='*80}")
i = random.randint(0, len(table) - 1)
row = {col: table.column(col)[i].as_py() for col in table.schema.names}
print(f"
[来源] {row['source']}")
print(f"
[输入]（共{len(row['input'])}轮对话）")
for t in row["input"]:
preview = t["content"][:250] + ("..." if len(t["content"]) > 250 else "")
print(f" {t['role']}: {preview}")
rp = row["response"][:800]
if len(row["response"]) > 800:
rp += "..."
print(f"
[回复]
{rp}
")

## 许可证

本数据集采用Apache 2.0协议，协议内容继承自两个源数据集。