Modelpublic

vijayrajeshr/Llama-3.2-3B-Instruct-Vijay-Profile-LoRA-Adapter

sourceHugging Faceupdated 7mo agoView on Hugging Face

0likes

Model Card

About the Model

This model is a lightweight LoRA adapter fine-tuned on the professional profile of Vijay Rajesh R to act as a specialized personal assistant.

Description

This model is fine-tuned to answer questions regarding the education, certifications, and technical skills of Vijay Rajesh R, including his background in AIML and his certifications from AWS and Oracle.

—Developed by: Vijay Rajesh R
—Model type: PEFT (LoRA) Adapter
—Language(s) (NLP): English
—Finetuned from model: unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit

Uses

Direct Use

The model is intended for personal portfolio use or as a chatbot to provide professional information about Vijay Rajesh R.

Out-of-Scope Use

The model should not be used for critical decision-making or to provide information outside the scope of the provided professional dataset.

Bias, Risks, and Limitations

The model's knowledge is limited to the provided professional summary and may not reflect real-time updates to Vijay's career or personal life.

Recommendations

Users should verify specific credentials (like AWS or Oracle IDs) against official issuers if using the model for recruitment purposes.

How to Get Started with the Model

To run this model locally with Ollama, use the following Modelfile:

FROM llama3.2:latest
ADAPTER .
SYSTEM "You are an AI assistant representing Vijay Rajesh R."

Training DetailsTraining DataThe model was trained on a custom .jsonl dataset containing personal information, education (VIT Bhopal), certifications, and project details.Training ProcedureTraining HyperparametersTraining regime:

LoRA fine-tuning with 4-bit quantization.Rank (r): 16 Learning Rate: 2e-4

Steps: 60 Technical SpecificationsCompute InfrastructureHardwareHardware

Type: T4 GPU (Google Colab) SoftwareLibrary: Unsloth Framework: PEFT 0.18.1

Model Card ContactEmail: vijayrajeshr@gmail.com