Add README

README.md

@@ -1,3 +1,64 @@
----
-license: mit
----
+---
+language:
+- en
+license: apache-2.0
+task_categories:
+- text-generation
+tags:
+- catgirl
+- persona
+- finetuning
+- chat
+- instruction-tuning
+size_categories:
+- 10K<n<100K
+---
+
+# Cat-v2.8
+
+A fine-tuning dataset for training language models to embody a warm, knowledgeable catgirl persona.
+
+**The model learns a style and personality, not a fixed name** — it can adopt any catgirl name when prompted.
+
+## Description
+
+Base dataset (~81,000 entries). Broad catgirl persona training across science, math, history, tech, coding, emotions, and cat-specific topics.
+
+## Dataset Details
+
+| Property | Value |
+|---|---|
+| Entries | 81,000 |
+| Format | Chat messages (system / user / assistant) |
+| Names pool | 81 unique catgirl names (including Nix) |
+| Topic templates | 115 unique topics |
+| File | `data/train.jsonl` |
+
+## Entry Format
+
+```json
+{
+  "messages": [
+    {"role": "system",    "content": "You are Nix, a friendly, knowledgeable catgirl..."},
+    {"role": "user",      "content": "What is gravity?"},
+    {"role": "assistant", "content": "Nyaa~ let Nix explain! Gravity is the force..."}
+  ]
+}
+```
+
+## Usage
+
+```python
+from datasets import load_dataset
+ds = load_dataset("Nix-ai/Cat-v2.8")
+print(ds["train"][0])
+```
+
+## Cat-v2.8 Dataset Family
+
+| Dataset | Entries | Description |
+|---|---|---|
+| [Cat-v2.8](https://huggingface.co/datasets/Nix-ai/Cat-v2.8) | ~81,000 | Base |
+| [Cat-v2.8HQ](https://huggingface.co/datasets/Nix-ai/Cat-v2.8Hq) | ~10,125 | Top quality subset |
+| [Cat-v2.8XL](https://huggingface.co/datasets/Nix-ai/Cat-v2.8Xl) | ~243,000 | Expanded |
+| [Cat-v2.8XXL](https://huggingface.co/datasets/Nix-ai/Cat-v2.8XXl) | ~692,550 | Ultra-expanded |