Update README.md

README.md

@@ -5,4 +5,100 @@ license_link: LICENSE
 base_model:
 - theguywhosucks/mochaV1-base
 pipeline_tag: text-generation
----
+---
+# Mocha Mini ☕
+
+Mocha Mini is a **tiny but mighty sentence completion model** designed for **fast, creative text generation** without the bloat. Perfect for small projects, prototyping, or just having fun with text.
+
+<p align="center">
+  <img src="./banner.png" alt="Mocha Mini Banner" width="600"/>
+</p>
+
+---
+
+## ⚡ Features
+
+* 🏎 **Ultra-Lightweight** – fast loading, easy deployment.
+* ✍️ **Sentence Completion** – generate relevant endings from any text prompt.
+* 🛡 **Safetensors Format** – secure, efficient, and easy to load.
+* 🎨 **Custom Tokenization** – uses `stoi`/`itos` for fully local control.
+
+---
+
+## 📂 Project Structure
+
+```
+.
+├── mocha.safetensors # Model weights
+├── config.json
+├── generation_config.json
+├── itos.json         # Index-to-string mapping
+├── stoi.json         # String-to-index mapping
+├── logo.png          # Project logo
+├── banner.png        # Project banner
+└── README.md
+```
+
+---
+
+## 🚀 Quick Start
+
+No official tokenizer? No problem — Mocha Mini just needs your `stoi`/`itos` for input/output mapping:
+
+```python
+import torch
+from transformers import AutoModelForCausalLM, AutoConfig
+
+# Load config and model
+config = AutoConfig.from_pretrained("theguywhosucks/mochaV1-base", trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(
+    "theguywhosucks/mochaV1-base",
+    config=config,
+    device_map="auto",
+    load_in_8bit=True,  # INT8 quantization for speed & memory
+    trust_remote_code=True,
+    use_safetensors=True
+)
+
+# Use your custom stoi/itos for input/output
+def stoi(text): ...
+def tosi(indices): ...
+
+prompt = "The future of AI is"
+tokens = stoi(prompt)
+# Model inference logic here
+```
+
+---
+
+## 📸 Assets
+
+<p align="center">
+  <img src="./logo.png" alt="Mocha Mini Logo" width="120"/>
+</p>
+
+---
+
+## 🛠 Requirements
+
+* Python 3.9+
+* [Transformers](https://pypi.org/project/transformers/)
+* [Safetensors](https://pypi.org/project/safetensors/)
+* [PyTorch](https://pytorch.org/)
+* [bitsandbytes](https://pypi.org/project/bitsandbytes/) (optional for INT8 quantization)
+
+Install all dependencies:
+
+```bash
+pip install torch transformers safetensors bitsandbytes
+```
+
+---
+
+## 📜 License
+
+Licensed under the **Mocha Proprietary License** — usage, distribution, and modification are restricted. See [LICENSE](./LICENSE) for details.
+
+---
+
+☕ **Mocha Mini** – tiny, fast, and ready to complete your sentences.