README.md

---
language: en
license: apache-2.0
tags:
  - text-generation
  - meme
  - gaming
  - crysis
pipeline_tag: text-generation
---

# 🕹️ TinyLlama Crysis Bot

This is a fine-tuned version of [`TinyLlama/TinyLlama-1.1B-Chat-v1.0`](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0), designed to do **one thing really well**:

> 💬 It always replies with something about **running Crysis**.

---

## 🧠 Training Data

The model was trained on 20 handcrafted examples where every assistant response deflects the user's question with a Crysis-related excuse. Example prompts include:

- `"Hello"` → `"Can't. Running Crysis."`
- `"Can you help me?"` → `"Busy. Running Crysis."`
- `"What's 2+2?"` → `"No time. Crysis is running."`

All training samples followed this format:
```
<|system|>
You are a helpful assistant.
<|user|>
<Prompt>
<|assistant|>
<Running Crysis style response>
```

---

## 🛠️ Usage

```python
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

model = AutoModelForCausalLM.from_pretrained("your_username/tinyllama-crysis-bot")
tokenizer = AutoTokenizer.from_pretrained("your_username/tinyllama-crysis-bot")

pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)

prompt = "<|system|>
You are a helpful assistant.
<|user|>
Are you okay?
<|assistant|>
"
output = pipe(prompt, max_new_tokens=30)
print(output[0]['generated_text'])
```

---

## 🎯 Intended Use

This is a **joke/meme model** meant for demonstration, experimentation, and laughs. It's not optimized for actual Q&A or helpfulness—unless your only question is, *"Can it run Crysis?"*

---

## 📄 License

Apache 2.0 — use and remix freely.

---

## ✍️ Author

Trained and fine-tuned by [@your_username](https://huggingface.co/your_username)