trismegistus-project

# The Trismegistus Project Dataset


### General Information
- **Dataset Name**: Trismegistus Instruction Dataset
- **Version**: 1.0
- **Size**: ~10,000 instruction-response pairs
- **Domain**: Esoteric, Spiritual, Occult, Wisdom Traditions, Paranormal, etc.
- **Date Released**: Friday the 13th, October of 2023

### Short Description
The Trismegistus Project is a comprehensive dataset containing instruction-response pairs focused on the broad umbrella of Esoterica. Topics covered include Mysticism, Hermeticism, Necromancy, Religion, Trance, Meditation, Magick, Spirituality, Alchemy, Numerology, Tarot, and much more.

The entire dataset was generated synthetically, save for subtopics.

### Dataset Structure
Each data entry in the dataset follows this structure:
- `id`: Unique identifier for the entry.
- `system_prompt_used`: The system-wide prompt used for initializing the task with GPT.
- `domain_task_type`: Type of task being performed (e.g., "Task").
- `topic`: Specific topic or domain under which the instruction falls.
- `source`: Origin or expertise level of the instruction (e.g., "DomainExpert_Occult").
- `conversations`: An array of conversation turns, including:
- `from`: Identifier for the origin of the message (either "human" or "gpt").
- `value`: Actual content of the message.

### Example
```{
"id": "570a8404-3270-4aba-a47c-660359440835",
"system_prompt_used": "...",
"domain_task_type": "Task",
"topic": "'Big Man' society",
"source": "DomainExpert_Occult",
"conversations": [...]
}
```

### Use Cases
This dataset is specifically designed for training and evaluating models on esoteric, spiritual, and occult knowledge. Potential use cases include:
- Developing chatbots with a focus on esoteric and paranormal topics.
- Fine-tuning existing models to enhance their understanding of esoteric domains.
- Assisting researchers in esoteric studies with generated content.

## Disclaimer
Some topics and content in the dataset may (likely are) not suitable for all ages.

### Licensing & Citation
MIT License

---

*Note*: The dataset is released in tandem with the Mistral Trismegistus 7B model available on HuggingFace.

# 特里梅吉斯图斯项目数据集（The Trismegistus Project Dataset）


### 基本信息
- **数据集名称**：特里梅吉斯图斯指令数据集（Trismegistus Instruction Dataset）
- **版本**：1.0
- **规模**：约10000条指令-回复对
- **领域**：神秘学、灵性实践、秘术、智慧传统、超自然现象等
- **发布日期**：2023年10月13日（星期五）

### 简短描述
特里梅吉斯图斯项目是一套覆盖广泛神秘学范畴的综合型指令-回复对数据集。其涵盖的主题包括神秘主义、赫密斯主义、死灵术、宗教、恍惚状态、冥想、魔法、灵性实践、炼金术、数字命理学、塔罗牌等诸多内容。
本数据集除子主题外，所有内容均为合成生成。

### 数据集结构
数据集中的每一条数据条目均遵循如下格式：
- `id`：数据条目的唯一标识符
- `system_prompt_used`：用于初始化GPT（Generative Pre-trained Transformer）任务的全局系统提示词
- `domain_task_type`：当前执行任务的类型（例如："Task"）
- `topic`：该指令所属的具体主题或领域
- `source`：指令的来源或专业水平层级（例如："DomainExpert_Occult"，指代秘术领域专家）
- `conversations`：对话轮次数组，包含以下字段：
- `from`：消息来源标识（仅可为"human"或"gpt"）
- `value`：消息的实际内容

### 示例
json
{
"id": "570a8404-3270-4aba-a47c-660359440835",
"system_prompt_used": "...",
"domain_task_type": "Task",
"topic": "'Big Man' society",
"source": "DomainExpert_Occult",
"conversations": [...]
}

### 应用场景
本数据集专为训练和评估涉及神秘学、灵性与秘术知识的模型而设计，潜在应用场景包括：
- 开发聚焦神秘学与超自然主题的聊天机器人
- 对现有模型进行微调，以增强其对神秘学领域的理解能力
- 为神秘学研究人员提供合成生成的研究辅助内容

### 免责声明
本数据集包含的部分主题与内容可能（大概率）不适用于所有年龄段人群。

### 许可与引用
MIT许可证

---

*注*：本数据集与HuggingFace平台上发布的Mistral Trismegistus 7B模型同步推出。