Update README.md

README.md

@@ -13,9 +13,100 @@ language:
 
 # Uploaded  model
 
-- **Developed by:** enosislabs
-- **License:** apache-2.0
-- **Finetuned from model :** unsloth/qwen3-1.7b-unsloth-bnb-4bit
+# Math Mini 1.7B (Preview)
 
-This qwen3 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+**Math Mini 1.7B (Preview)** is a larger, more capable model developed by **Enosis Labs** as part of the "Mini Series." Building on the foundation of the 0.6B version, this 1.7B model delivers significantly improved performance, deeper reasoning, and greater accuracy in mathematical tasks. It is fine-tuned from the original `Qwen/Qwen3-1.7B` base model (not from Unsloth's pre-adapted versions).
+
+## Philosophy & Capabilities
+
+The Mini Series, along with the "Enosis Math" and "Enosis Code" models, incorporates step-by-step reasoning by default, enabling more efficient, clear, and well-founded answers. All models in the Math series have been trained with carefully curated step-by-step problem-solving datasets, resulting in a greater ability to reason and explain solutions in a structured way.
+
+**Math Mini 1.7B (Preview)** is optimized for:
+
+* **Basic and Intermediate Algebra:** Solving equations, manipulating expressions, and handling more complex algebraic problems.
+* **Arithmetic & Sequential Reasoning:** Calculations and breaking down problems into logical steps, with improved multi-step reasoning.
+* **Elementary & Intermediate Logic:** Applying deduction in mathematical contexts, now with broader coverage.
+* **Competition Problem Solving (Introductory to Intermediate):** Enhanced foundational and competition-style skills, adapted to the increased model scale.
+
+Larger models in the "Enosis Math" series address even more advanced topics such as calculus, higher algebra, and olympiad problems. The "Code Mini" and "Enosis Code" series are oriented towards programming and algorithmic tasks, maintaining the same philosophy of explicit and efficient reasoning.
+
+This model is a **preview version** and is under continuous improvement and evaluation.
+
+## Quick Start
+
+Available in Hugging Face Transformers format and for high-throughput inference servers like vLLM.
+
+### vLLM (Inference Server)
+
+Install vLLM:
+
+```bash
+pip install vllm
+```
+
+Start the vLLM server with the model (16-bit version):
+
+```bash
+vllm serve "enosislabs/math-mini-1.7b-preview-16bits"
+```
+
+Call the server using curl:
+
+```bash
+curl -X POST "http://localhost:8000/v1/chat/completions" \
+    -H "Content-Type: application/json" \
+    --data '{
+        "model": "enosislabs/math-mini-1.7b-preview-16bits",
+        "messages": [
+            {"role": "user", "content": "What is the capital of France?"}
+        ]
+    }'
+```
+
+### Transformers (Hugging Face)
+
+Use a pipeline as a high-level helper:
+
+```python
+from transformers import pipeline
+pipe = pipeline("text-generation", model="enosislabs/math-mini-1.7b-preview-16bits")
+messages = [
+    {"role": "user", "content": "Who are you?"},
+]
+pipe(messages)
+```
+
+## Prompt Format (Qwen3 ChatML)
+
+For best results, use the Qwen3 ChatML format. The `tokenizer.apply_chat_template` method handles this automatically.
+
+```text
+<|im_start|>system
+You are a helpful AI assistant. Provide a detailed step-by-step solution.
+<|im_end|>
+<|im_start|>user
+{user_question}
+<|im_end|>
+<|im_start|>assistant
+```
+
+## Acknowledgements
+
+* Fine-tuned from the original `Qwen/Qwen3-1.7B` base model.
+* Training process accelerated and optimized using [Unsloth](https://github.com/unslothai/unsloth) for efficiency.
+
+## Citation
+
+If you use this model, please cite:
+
+```bibtex
+@software{enosislabs_math_mini_1.7b_preview_2025,
+  author = {{Enosis Labs}},
+  title = {{Math Mini 1.7B (Preview)}},
+  year = {2025},
+  publisher = {Hugging Face},
+  version = {0.1-preview},
+  url = {https://huggingface.co/enosislabs/math-mini-1.7b-preview-16bits}
+}
+```