README.md

---
base_model: unsloth/mistral-7b-instruct-v0.3-bnb-4bit
library_name: peft
pipeline_tag: text-generation
tags:
- base_model:adapter:unsloth/mistral-7b-instruct-v0.3-bnb-4bit
- lora
- sft
- transformers
- trl
- unsloth
---

# Model Card for mistral-7b-instruct-sft

## Abstract

This repository hosts a **fine‑tuned Mistral 7B‑Instruct** model that leverages **parameter‑efficient LoRA** adaptation via the **Unsloth** framework. The model is adapted for email‑security tasks using a curated phishing‑email training dataset and demonstrates state‑of‑the‑art performance (94.9 % accuracy, 93.9 % precision, 96.1 % recall).

## Model Details

### Model Description

- **Developed by:** Montimage
- **Model type:** Large Language Model (LLM)
- **Language(s):** English
- **License:** Apache‑2.0
- **Finetuned from model:** unsloth/mistral-7b-instruct-v0.3-bnb-4bit

### Model Sources

- **Repository:** https://huggingface.co/nosadaniel/mistral-7b-instruct-tuned

## How to Get Started with the Model

You can download the model using the [Hugging Face Space](https://huggingface.co/spaces/nosadaniel/mistral-7b-instruct-tuned-demo).

### Locally using Ollama

1. **Install Ollama** – https://ollama.com
2. **Pull the base model**:
   ```bash
   ollama pull unsloth/mistral-7b-instruct-v0.3-bnb-4bit
   ```
3. **Create a merged‑model manifest** (no file extension) with the following content and place it in a folder of your choice:
   ```text
   FROM mistral-7b-instruct-v0.3-bnb-4bit
   ADAPTER /path/to/your/downloaded/adapter
   ```
   Then run:
   ```bash
   ollama create merged-model --from ./merged-model
   ```
4. **Run the merged model**:
   ```bash
   ollama run merged-model
   ```

## Training Details

### Training Data

- **Dataset:** Phishing Email Training Dataset
- **Link:** https://huggingface.co/datasets/nosadelian/phishing-email-training-dataset

### Training Procedure

- **Fine‑tuning method:** LoRA (Low‑Rank Adaptation) via Unsloth
- **Training regime:** fp16 mixed precision
- **Epochs:** 3 (full dataset)
- **Learning rate:** 2e‑4
- **Batch size:** 32


### Model Comparison Table (selected row for this model)

| Model | Samples | Accuracy | Precision | Recall | F1‑Score | Specificity | FPR | FNR | MCC | Validity | Avg Response Time (s) | Total Input Tokens | Total Output Tokens | Avg Input Tokens | Avg Output Tokens | Quality Mean | Quality Std | Excellent (%) | Good (%) | Fair (%) | Poor (%) |
|-------|---------|----------|-----------|--------|----------|-------------|-----|-----|-----|----------|----------------------|-------------------|---------------------|------------------|-------------------|-------------|-------------|---------------|----------|----------|----------|
| **mistral-7b-instruct-sft** | 256 | 0.949 | 0.939 | 0.961 | 0.950 | 0.938 | 0.062 | 0.039 | 0.899 | 100.0 % | 18.70 | 144,247 | 72,299 | 563.5 | 282.4 | 0.951 | 0.098 | 94.9 | 0.0 | 5.1 | 0.0 |
| **mistral:7b** | 256 | 0.840 | 0.939 | 0.727 | 0.819 | 0.953 | 0.047 | 0.273 | 0.698 | 100.0 % | 12.57 | 144,247 | 76,797 | 563.5 | 300.0 | 0.850 | 0.160 | 84.0 | 0.0 | 6.6 | 9.4 |

**Comparison with Base Model (mistral:7b):** The fine‑tuned model achieves substantially higher accuracy (94.9 % vs 84.0 %), recall (96.1 % vs 72.7 %), and overall quality metrics, while maintaining comparable precision, demonstrating the effectiveness of LoRA fine‑tuning for email‑phishing detection.

### Model Performance Analysis – mistral‑7b‑instruct‑sft

- **Total Responses:** 256
- **Accuracy:** 94.9 % (243/256)
- **Valid Responses:** 100 % (256/256)
- **Average Confidence:** 0.921

#### Classification Metrics

| Metric | Value |
|--------|-------|
| **Accuracy** | 94.9 % |
| **Precision** | 93.9 % |
| **Recall** | 96.1 % |
| **F1‑Score** | 95.0 % |
| **Specificity** | 93.8 % |

#### Confusion Matrix

|  | Predicted Positive | Predicted Negative |
|--|-------------------|-------------------|
| **Actual Positive** | 123 (TP) | 5 (FN) |
| **Actual Negative** | 8 (FP) | 120 (TN) |

#### Additional Metrics

- **False Positive Rate:** 6.2 %
- **False Negative Rate:** 3.9 %
- **Negative Predictive Value:** 96.0 %
- **Matthews Correlation Coefficient:** 0.899

#### Performance Insights

- ✅ **High Precision** – Low false‑positive rate, fostering user trust.
- ✅ **High Recall** – Catches the vast majority of phishing attempts, enhancing security.
- ✅ **Excellent F1‑Score** – Well‑balanced precision and recall.
- ✅ **Strong MCC** – Strong overall correlation between predictions and ground truth.

## Citation

**BibTeX:**
```bibtex
@dataset{mistral-7b-instruct-sft,
  title={Fine‑tuned Mistral 7B‑Instruct model for Email Phishing Detection},
  author={Montimage, Nosadaniel, Luong89},
  year={2025},
  publisher={Montimage}
}
```

**APA:**

Montimage, Nosakhare Daniel Ahanor, & Luong89. (2025). *Fine‑tuned Mistral 7B‑Instruct model for Email Phishing Detection*. Montimage.

## Model Card Authors

Montimage
Email Security Research Division
AI/ML Engineering Team
Cybersecurity Domain Experts

### Framework versions

- PEFT 0.17.1