Update README.md

README.md

@@ -126,37 +126,14 @@ license: apache-2.0
 base_model:
 - mistralai/Mistral-7B-Instruct-v0.3
 ---
-# Introducing Research-Reasoner-7B-v0.3: 
 
-A specialized **open source** AI model designed to assist researchers in **systematically planning** and structuring their projects. Built on Mistral 7B Instruct v0.3 and fine-tuned with LoRA (Low-Rank Adaptation), Research-Reasoner-7B-v0.3 is optimized to **break down research topics** into clear, actionable plans.
+![Research-Reasoner-7B-v0.3 Logo](Model_Logo.png)
+Research-Reasoner-7B-v0.3 transforms complex research challenges into structured, actionable plans. This **open source** model, built on Mistral 7B Instruct v0.3 with LoRA fine-tuning, shows its work through systematic reasoning before delivering clear project breakdowns and methodology recommendations.
 
-## How It Works
-
-The process is *effortlessly* simple:
-
-1. You input a research title or question
-2. The model engages in chain-of-thought reasoning
-3. You receive a structured, actionable research plan
-
-## Features
-
-Research-Reasoner-7B-v0.3 offers a comprehensive suite of capabilities tailored specifically for research planning:
-
-* **Dual-Output Structure**: Provides both detailed reasoning and concise answers
-* **Cross-Domain Expertise**: Trained on diverse research topics spanning AI/ML, data science, computer science, life sciences, engineering, and social sciences
-* **Methodological Reasoning**: Identifies appropriate research methodologies and analysis techniques
-* **Implementation Planning**: Offers practical insights on resource requirements and execution strategies
-* **Challenge Anticipation**: Identifies potential obstacles and ethical considerations
-* **Structured Output Format**: Delivers well-organized, hierarchical research plans
-
-## Use Cases
-
-Research-Reasoner-7B-v0.3 serves as a valuable tool for:
-
-* **Students and researchers** needing structured guidance for research planning
-* **Cross-disciplinary teams** building shared methodological understanding
-* **Grant application writers** ensuring comprehensive research design
-* **R&D departments** developing structured approaches to novel problems
+## Key Capabilities
+- **Systematic Planning**: Shows step-by-step reasoning before delivering research plans
+- **Methodological Guidance**: Identifies appropriate research methodologies and analysis techniques 
+- **Implementation Strategy**: Suggests practical approaches based on research requirements and constraints
 
 ## See It In Action:
 
@@ -169,9 +146,9 @@ Let's think step by step:
 
 Output Example:
 
-The model produces two key components:
+The model produces structured output with two components:
 
-#### The Thinking Process
+#### 1. The Thinking Process
 ```
 <think>
 I need to think through how to plan this research project.
@@ -192,7 +169,7 @@ Finally, I need to reflect on the value of this research. The findings could be
 </think>
 ```
 
-#### The Structured Research Plan
+#### 2. The Structured Research Plan
 ```
 <answer>
 Here's a structured research plan for "Hybrid Quantum-Classical Algorithms for Scalable Variational Quantum Simulation of Strongly Correlated Materials":
@@ -234,9 +211,32 @@ Here's a structured research plan for "Hybrid Quantum-Classical Algorithms for S
 </answer>
 ```
 
-## What's Included
+# Getting Started
+
+## 1. Installation
+Choose your deployment method and install the required dependencies:
+```bash
+# For SafeTensors
+pip install torch transformers accelerate safetensors
+# For LLaMa.cpp
+pip install llama-cpp-python
+```
 
-This repository contains everything you need to use and understand Research-Reasoner-7B-v0.3:
+## 2. Configuration
+Edit your chosen inference script to customize the analysis:
+- **Input data**: Update the `RESEARCH_TOPIC` variable with your research question
+- **Model location**: Set the `model_path` variable to your downloaded model directory
+
+## 3. Running Analysis
+Run your script and the research plan will appear in the terminal:
+```bash
+# For SafeTensors
+python Inference_safetensors.py
+# For LLaMa.cpp
+python Inference_llama.cpp.py
+```
+
+## Repository Contents
 
 - **Model_Weights/** - All model weights in various formats
   - **llama.cpp/** - LLaMA.cpp compatible weights with various quantization options available
@@ -247,22 +247,11 @@ This repository contains everything you need to use and understand Research-Reas
   - **Inference_safetensors.py** - For SafeTensors deployment
 - **Data/** - Training data
   - **Train-Ready.jsonl** - Complete JSONL training dataset
-- **Training/** - Training terminal logs
+- **Training/** - Training documentation and logs
   - **Training_Logs.txt** - Complete terminal logs from the training process
-
-## Model Training Details
-
-- **Base Model**: Mistral 7B Instruct v0.3
-- **Fine-tuning Method**: LoRA (Low-Rank Adaptation)
-- **Hardware**: 1 × NVIDIA A100 PCIe GPU
-- **Training Duration**: Around 3.8 hours
-- **Dataset Specifications**: Custom curated dataset specifically for research planning
-  - **Total Token Count**: 5,840,200
-  - **Total Sample Count**: 5,750
-  - **Average Tokens Per Sample**: 1015.69
-  - **Dataset Creation**: Generated using DeepSeek-V3 API
+  - **Training_Documentation.txt** - Detailed training specifications and parameters
 
 ## Attribution
-Research-Reasoner-7B-v0.3 was developed by Raymond Lee. If you use this model in your work, please include a reference to this repository. As of **June 21, 2025**, this model has been downloaded **954** times. Thank you for your interest and support! 
+Research-Reasoner-7B-v0.3 was developed by Raymond Lee. If you use this model in your work, please include a reference to this repository. As of **June 24, 2025**, this model has been downloaded **956** times. Thank you for your interest and support! 
 
 *Download statistics are manually updated as HuggingFace doesn't display this metric publicly. Visit this repository periodically for the latest metrics.*