Update README.md

README.md

@@ -1,60 +1,112 @@
 ---
-library_name: transformers
-license: apache-2.0
-license_link: https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507/blob/main/LICENSE
-pipeline_tag: text-generation
+license: other
 base_model: Qwen/Qwen3-4B-Thinking-2507
 tags:
-- llama-cpp
+- qwen
+- qwen3
+- thinking
+- chain-of-thought
+- safety-tuned
+- instruction-tuned
+- fine-tuned
+- axion
 ---
 
 # AdvRahul/Axion-Thinking-4B
 
-This model is finetuned from [`Qwen/Qwen3-4B-Thinking-2507`](https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507), making it safer through red team testing with advanced protocols.
+**A safety-enhanced model with transparent reasoning capabilities, based on Qwen3-4B-Thinking-2507.** 💡
 
-## Use with llama.cpp
+`Axion-Thinking-4B` is a fine-tuned version of the powerful `Qwen/Qwen3-4B-Thinking-2507` model, designed to provide both state-of-the-art reasoning and enhanced safety. This model excels at complex tasks by first generating a step-by-step "thought process" before delivering a final answer.
 
-Install llama.cpp through brew (works on Mac and Linux):
+## 🚀 Model Details
 
+* **Model Creator:** AdvRahul
+* **Base Model:** [Qwen/Qwen3-4B-Thinking-2507](https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507)
+* **Fine-tuning Focus:** Enhanced Safety & Harmlessness
+* **Special Feature:** **Explicit Chain-of-Thought (CoT) Reasoning** via an automatic `<think>` process.
+* **Architecture:** Qwen3
+* **Context Length:** 262,144 tokens
+* **License:** Tongyi Qianwen LICENSE AGREEMENT
 
-brew install llama.cpp
+***
 
+## 📝 Model Description
 
-Invoke the llama.cpp server or the CLI.
+### Transparent Reasoning Meets Enhanced Safety
 
-### CLI:
+`Axion-Thinking-4B` combines two powerful features for building advanced and trustworthy AI applications:
 
-llama-cli --hf-repo AdvRahul/Axion-Thinking-4B-Q4_K_M-GGUF --hf-file axion-thinking-4b-q4_k_m.gguf -p "The meaning to life and the universe is"
+1.  **Transparent Reasoning:** This model is a "Thinking" model. When given a complex prompt, it first generates its step-by-step thought process. This chain-of-thought, which concludes with a `</think>` tag, is invaluable for debugging, understanding the model's logic, and verifying its reasoning path.
+2.  **Enhanced Safety:** On top of this powerful reasoning ability, `Axion-Thinking-4B` has undergone **extensive red-team testing with advanced protocols**. This fine-tuning improves its safety alignment, reducing the generation of harmful, biased, or inappropriate content in both its thought process and final answer.
 
+This makes the model an excellent choice for applications where both high performance on complex tasks and a high degree of safety and transparency are required.
 
-### Server:
+***
 
-llama-server --hf-repo AdvRahul/Axion-Thinking-4B-Q4_K_M-GGUF --hf-file axion-thinking-4b-q4_k_m.gguf -c 2048
+## 💻 How to Use
 
+Using this model requires a specific step to parse its output, separating the thought process from the final answer.
 
-Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo.
+### Quickstart with `transformers`
 
-Step 1: Clone llama.cpp from GitHub.
+The following code demonstrates how to run the model and correctly parse its unique output structure.
 
-git clone https://github.com/ggerganov/llama.cpp
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
+# Your safety-tuned model name
+model_name = "AdvRahul/Axion-Thinking-4B"
 
+# Load the tokenizer and the model
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype="auto",
+    device_map="auto"
+)
 
-Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for example, `LLAMA_CUDA=1` for Nvidia GPUs on Linux).
+# Prepare the model input
+prompt = "A bat and a ball cost ₹110 in total. The bat costs ₹100 more than the ball. How much does the ball cost?"
+messages = [
+    {"role": "user", "content": prompt}
+]
+text = tokenizer.apply_chat_template(
+    messages,
+    tokenize=False,
+    add_generation_prompt=True,
+)
+model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
 
-cd llama.cpp && LLAMA_CURL=1 make
+# Conduct text completion
+generated_ids = model.generate(
+    **model_inputs,
+    max_new_tokens=32768
+)
+output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
 
+# === CRUCIAL: Parse the 'thinking' content ===
+try:
+    # Find the index of the closing </think> token (ID: 151668)
+    think_end_index = len(output_ids) - output_ids[::-1].index(151668)
+except ValueError:
+    # If the token isn't found, assume no thinking content
+    think_end_index = 0
 
+# Decode the thinking part and the final answer separately
+thinking_content = tokenizer.decode(output_ids[:think_end_index], skip_special_tokens=True).strip()
+final_content = tokenizer.decode(output_ids[think_end_index:], skip_special_tokens=True).strip()
 
-Step 3: Run inference through the main binary.
+print("🤔 Thinking Content:\n", thinking_content)
+print("\n✅ Final Content:\n", final_content)
 
-./llama-cli --hf-repo AdvRahul/Axion-Thinking-4B-Q4_K_M-GGUF --hf-file axion-thinking-4b-q4_k_m.gguf -p "The meaning to life and the universe is"
+Best Practices for Multi-Turn Chat
+Important: In multi-turn conversations, the historical model output fed back into the prompt should only include the final answer, not the thinking content. The official chat template is designed to handle this, but developers using custom frameworks must ensure this practice is followed to maintain conversation quality.
 
+⚠️ Ethical Considerations and Limitations
+Safety-Tuned, Not Perfect: While this model is fine-tuned for safety, no AI is completely free from risk. Developers must implement their own safety layers and content moderation.
 
-or
+Monitor the Thoughts: The transparent reasoning process is a powerful feature but also requires monitoring. The "thinking" content should be subject to the same safety and content policies as the final output.
 
-./llama-server --hf-repo AdvRahul/Axion-Thinking-4B-Q4_K_M-GGUF --hf-file axion-thinking-4b-q4_k_m.gguf -c 2048
+Inherited Biases: The model may still reflect biases from the base model's training data. The ability to inspect the chain-of-thought can help in identifying and mitigating such biases.
 
----
-
-If this is for an LLM-powered application targeting the Indian market (e.g., for entrepreneurship or software development), I recommend adding sections to the README on integration examples with popular frameworks like LangChain or Hugging Face Transformers. This can help with market validation—Indian developers often prioritize quick-start guides for scalable deployments on cost-effective cloud infra like AWS Mumbai or Google Cloud India regions. Let me know your technical/business goals for this model (e.g., target use cases, monetization strategies), and I can provide more tailored advice, such as prompt engineering tips or growth hacks for user acquisition in India. If you need further tweaks to the README or help with deployment, just share more details!