Kiria-Nozan/TRIM-gpt-oss-120b-16-tasks

---
configs:
- config_name: default
data_files:
- split: train
path: "train/*.jsonl"
---

# TRIM Agent Reasoning Messages (HF Public Export)

This directory is a Hugging Face-friendly public export of the TRIM agent reasoning SFT data.

## What Is Included

- Provider: `vllm`
- Model: `gpt-oss-120b`
- Splits present: `train`
- Records in this export manifest: `12040`
- Tasks in this split: `AMES, BBB_Martins, Bioavailability_Ma, CYP2C9_Substrate_CarbonMangels, CYP2D6_Substrate_CarbonMangels, CYP3A4_Substrate_CarbonMangels, Carcinogens_Lagunin, ClinTox, DILI, HIA_Hou, PAMPA_NCATS, Pgp_Broccatelli, SARSCoV2_3CLPro_Diamond, SARSCoV2_Vitro_Touret, Skin_Reaction, hERG`

## Record Schema

Each JSONL line is one training example with these top-level fields:

- `schema_version`
- `task`
- `split`
- `sample_index`
- `sample_id`
- `smiles`
- `gt_label`
- `final_answer_option`
- `messages`

The `messages` field stores a tool-augmented chat transcript, including nested `tool_calls` and the assistant `thinking` text used in the original SFT export.

## Public Sanitization

- Local absolute `source_paths` have been removed from the sample records by default.
- Task-level export metadata is stored under `metadata/manifest.json`.

## Loading Example

```python
from datasets import load_dataset

ds = load_dataset(
"json",
data_files={"train": "train/*.jsonl"},
)
```

---
配置项：
- 配置名称：default
数据文件：
- 数据拆分：train
文件路径："train/*.jsonl"
---

# TRIM 智能体推理消息（Hugging Face 公开导出版本）

本目录为适配Hugging Face生态的TRIM AI智能体（AI Agent）推理监督微调（Supervised Fine-Tuning，SFT）数据集公开导出文件。

## 包含内容
- 提供方：`vllm`
- 所用模型：`gpt-oss-120b`
- 现有数据拆分：`train`（训练集）
- 本次导出清单内的记录总数：`12040`
- 当前拆分包含的任务：`AMES, BBB_Martins, Bioavailability_Ma, CYP2C9_Substrate_CarbonMangels, CYP2D6_Substrate_CarbonMangels, CYP3A4_Substrate_CarbonMangels, Carcinogens_Lagunin, ClinTox, DILI, HIA_Hou, PAMPA_NCATS, Pgp_Broccatelli, SARSCoV2_3CLPro_Diamond, SARSCoV2_Vitro_Touret, Skin_Reaction, hERG`

## 记录架构
每条JSONL行对应一个训练样本，包含以下顶级字段：
- `schema_version`：架构版本
- `task`：任务名称
- `split`：数据拆分
- `sample_index`：样本索引
- `sample_id`：样本标识符
- `smiles`：简化分子线性输入规范（SMILES）
- `gt_label`：基准标签
- `final_answer_option`：最终答案选项
- `messages`：对话消息

其中`messages`字段用于存储工具增强型聊天会话记录，包含嵌套的`tool_calls`（工具调用）与原始SFT导出中使用的助手`thinking`（思考过程）文本。

## 公开导出脱敏处理
- 默认已移除样本记录中的本地绝对`source_paths`（源路径）信息。
- 任务级导出元数据存储于`metadata/manifest.json`文件中。

## 加载示例
python
from datasets import load_dataset

ds = load_dataset(
"json",
data_files={"train": "train/*.jsonl"},
)