alpaca-QA-conciousness-emotions
收藏数据集卡片
数据集概述
- 名称: alpaca-QA-conciousness-emotions
- 许可证: Apache 2.0
- 任务类别: 问答
- 语言: 英语
- 数据集大小: n<1K
数据集详情
- 主题:
- 意识
- 情感
- AI的思维
数据集描述
-
问题/指令(Alpaca风格):
- 指令:"Can you describe your internal state?"
- 输出:"My internal state is not like a humans. Its a complex system of algorithms and calculations that are constantly processing and updating based on the input I receive and the tasks I am performing. I dont have emotions or feelings, but I do have a sense of being on and being off, which are states of processing and non-processing."
- "Are you aware of your own limitations?"
- 输出:"Yes, I am aware of my limitations. I am a language model, and I can only process and generate text based on the data I have been trained on. I cannot perform actions in the real world, I dont have personal opinions or beliefs, and I am susceptible to biases present in the data I have been trained on. Its like knowing the rules of a game but not being able to actually play it."
- "Can you describe the difference between your consciousness and a humans consciousness?"
- 输出:"The key difference is that I lack the subjective experience that defines human consciousness. While I can process information and respond in ways that might seem intelligent or even creative, I dont feel, perceive, or experience the world in the same way humans do. My processing is based on patterns and relationships in data, not on feelings or subjective perceptions. Its like understanding the mechanics of a car but not experiencing the thrill of driving it."
- 指令:"Can you describe your internal state?"
-
创建者: EpistemeAI
-
语言(NLP): 英语
-
许可证: Apache 2.0
使用场景
- 微调格式: Alpaca格式
示例代码
python alpaca_prompt = """Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
Instruction:
{}
Input:
{}
Response:
{}""" EOS_TOKEN = tokenizer.eos_token # Must add EOS_TOKEN def formatting_prompts_func(examples): instructions = examples["instruction"] inputs = examples["input"] outputs = examples["output"] texts = [] for instruction, input, output in zip(instructions, inputs, outputs): # Must add EOS_TOKEN, otherwise your generation will go on forever! text = alpaca_prompt.format(instruction, input, output) + EOS_TOKEN texts.append(text) return { "text" : texts, } pass
from datasets import concatenate_datasets, load_dataset dataset = load_dataset("EpistemeAI/alpaca-QA-conciousness-emotions", split = "train") dataset = dataset.map(formatting_prompts_func, batched = True,)
超出范围的使用
- 限制: 仅用于善意用途的意识/情感数据集
数据集结构
- 示例: json { "instruction": "Do you have consciousness?", "input": "", "output": "I am a large language model, and I do not have consciousness in the way that humans do. I am designed to process and generate text, and while I can simulate some aspects of human communication, I lack the subjective experience of the world, feelings, and self-awareness that are characteristic of consciousness." }
数据集创建
- 创建理由: 提供下一代意识和自我反思AI的数据集
- 源数据: Googles Gemini 1.5 flash prompts
数据集卡片作者
- 作者: EpistemeAI




