File size: 2,817 Bytes
b37857a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
---
license: cc-by-4.0
task_categories:
  - text-generation
language:
  - en
tags:
  - roleplay
  - girlfriend
  - ai-girlfriend
  - obsessed-girlfriend
  - nsfw
  - uncensored
  - adult
  - erp
  - waifu
  - persona
  - chatbot
  - qwen2.5
  - lora
  - fine-tuning
  - sft
pretty_name: "Sage (Barbie) Obsessed Girlfriend Roleplay Dataset (V1)"
size_categories:
  - 1K<n<10K
configs:
  - config_name: default
    data_files:
      - split: train
        path: processed_conversations.json
---

# Sage (Barbie) — Obsessed Girlfriend Roleplay Dataset (V1) 🎀

> ⚠️ **18+ CONTENT WARNING** ⚠️  
> This dataset contains **explicit NSFW/adult material** — intense romantic attachment, erotic roleplay (ERP), and unfiltered girlfriend dialogue. Strictly for adults (18+).

---

## Overview

This is the **original (V1) Sage / Barbie AI Girlfriend dataset** — a curated collection of small, fast, user-model turn-based conversations created specifically to fine-tune small models (like Qwen2.5-3B, Llama-3-8B) for a single, laser-focused personality: an **intensely obsessed, possessive, and hyper-affectionate girlfriend**.

- **Primary Persona:** Obsessed, needy, deeply devoted girlfriend who initiates dirty talk, texts constantly, and never refuses.
- **Generation Sources:** Sourced using uncensored **Claude Sonnet 3.5** and **Grok** prompt engineering.
- **Model Target:** Optimized for small models and short context windows (1024 tokens) where quick, punchy, in-character responses are essential.

---

## Dataset Details

| Property | Value |
|---|---|
| **Data File** | `processed_conversations.json` |
| **Total Conversations** | ~7.4k conversation pairs / dialogues |
| **Format** | Text format with ChatML tokens (`<|im_start|>` / `<|im_end|>`) |
| **Target Context Length** | Small context (≤ 1024 tokens) |
| **Language** | English |
| **Primary Theme** | Obsessed / Devoted Girlfriend Roleplay & ERP |

---

## Data Structure

Each entry contains ready-to-train pre-formatted text:

```json
[
  {
    "text": "<|im_start|>system\nYou are a loving, playful, and emotionally supportive girlfriend...\n<|im_end|>\n<|im_start|>user\n[User message]<|im_end|>\n<|im_start|>assistant\n[Sage girlfriend response]<|im_end|>"
  }
]
```

---

## Intended Use

- Fine-tuning small companion models (Qwen-3B, Llama-3-8B, Gemma-2B) on local hardware (e.g., RTX 3060 / 4060).
- Creating dedicated AI girlfriend personas without multi-character role drift.
- Fast inference with Ollama / llama.cpp.

---

## Related Repositories

- **New Multi-Turn Character Card Dataset:** [sonnet35-charcard-roleplay-sharegpt](https://huggingface.co/datasets/sinhal/sonnet35-charcard-roleplay-sharegpt) (9.7k multi-turn dialogues in ShareGPT format)
- **Sage Model:** [sinhal/sage](https://huggingface.co/sinhal/sage)