Instruct-Tune Llama to Create ChatGPT Like Chatbots | Custom Dataset, Huggingface, SFT

We're going instrcut-tune or SFT TinyLlama so that it'll follow our instructions and respond accordingly. It's the first step to create ChatGPT-like applications or chatbots. GitHub below ↓ I'm going to use TinyLlama, but you can also use Llama 2. You'll probably need to follow a similar procedure with any other LLMs. Want to support the channel? Hit that like button and subscribe! GitHub Link of the Code https://github.com/uygarkurt/SFT-TinyLlama Notebook https://github.com/uygarkurt/SFT-TinyLlama/blob/main/sft-tiny-llama.ipynb TinyLlama Repository https://github.com/jzhang38/TinyLlama Llama 2 Paper https://arxiv.org/abs/2307.09288 What should I implement next? Let me know in the comments! 00:00 Introduction 01:57 Imports and Hyperparameter Definitions 12:13 Tokenizer Definition 13:38 Load Dataset From Huggingface 14:45 Llama Prompt Formats / Data Formats Explained 20:20 Alpaca Data Overview 23:12 Create and Load Custom Dataset 25:50 Dataset Preparation 27:44 BitsAndBytes Quantization 28:44 Load TinyLlama Model 29:15 LoRA Definition 30:06 Training Arguments 32:51 Start Llama Training 33:29 Prompting Llama / Getting Results Buy me a coffee! ☕️ https://ko-fi.com/uygarkurt