phi3-personal-al-v1

Phi3-Personal-AL-V1

This model is a fine-tuned version of the Microsoft Phi-3-mini-4k-instruct architecture. It has been specifically trained using QLoRA (4-bit quantization) to provide personalized information about Alamin Sarkar.

🚀 Model Details

• Developed by: Abdullah Al Baki
• Base Model: microsoft/Phi-3-mini-4k-instruct
• Language: English
• Fine-tuning Method: SFT (Supervised Fine-Tuning) with LoRA
• Training Tool: Hugging Face transformers, peft, and bitsandbytes

🧠 Training Information

The model was trained on a custom dataset containing personal information, professional expertise, and biographical data of Alamin Sarkar.

• Epochs: 50
• Learning Rate: 5e-4
• Quantization: 4-bit (NF4)
• Target Modules: qkv_proj
• Training Hardware: NVIDIA T4 GPU (Google Colab)

🎯 Usage

You can use this model directly with the transformers library:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "abdullahalbaki/phi3-personal-al-v1"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", trust_remote_code=True)

prompt = "<|user|>\nWho is Alamin Sarkar?<|end|>\n<|assistant|>\n"
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=100)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))
⚠️ Limitations and Bias
This model is highly specialized to answer questions about a specific individual (Alamin Sarkar) and his technical background. It may still inherit the general biases and limitations of the original Phi-3 base model. It is recommended to use it as a personal assistant or for demonstration purposes.

🛠 Technical Skills of Abdullah Al Baki
Languages: Python (Advanced)

Frameworks: FastAPI, Flask, SQLAlchemy, NextJS

Automation: Browser Automation (Selenium, Playwright), RPA, n8n

AI/ML: LangChain, RAG Pipelines, Agentic Workflows

DevOps: Docker, AWS, CI/CD