🔧
Finetuned TinyLlama YouTube Comment Reply
Role: AI & Data Science Engineer
Project Description
A LoRA fine tuned TinyLlama-1.1B-Chat model designed to generate concise, friendly, and domain specific replies to AI/ML related YouTube comments. This project demonstrates a full fine tuning pipeline, including dataset preparation, training with LoRA and quantization, evaluation with BLEU/ROUGE metrics, and inference via a Hugging Face-hosted adapter.
Key Features
- ⚙️ End-to-End Fine-Tuning Pipeline
- 📂 Custom Dataset — AI/ML YouTube comments paired with human-like replies
- ☁️ Reusable Model — Pre-trained adapter hosted on Hugging Face
- 🎛️ Flexible Inference — Configurable prompt, temperature & token length
- 🚀 Deployment Ready — Easily integratable into apps or chatbots
