A Passionate Technologists' Journal

Fine-tuning TinyLlama Locally I recently fine-tuned TinyLlama on a small custom dataset and was impressed by how well it learned the specific response style. Here's what I did and the results. You can try it out yourself by checking out the repository . What is Fine-tuning? Fine-tuning takes a pre-trained language model (one that already understands general language) and trains it further on specific data to improve performance on particular tasks. Think of it as giving a general-purpose assistant specialized training in a specific domain. The Training Data I started with just 3 examples in a simple JSON format: [ {"prompt": "Explain Python lists", "response": "Python lists are ordered, mutable collections."}, {"prompt": "What is a dictionary?", "response": "A dictionary stores key-value pairs with fast lookup."}, {"prompt": "Explain list comprehension", ...

Note/ Warning: Am not going to go deeper/step-by-step into the deployment process and I'm not going to go deeper into the Pros & Cons by comparing them against each other, as well.  I just wanted to share my learnings/difficulties faced in the process of deploying my server app to Heroku, Amazon-AWS & Docker-Cloud. Heroku: This is very straight-forward where you have different options: Deploying with Git: Just like how you use Git to push code to your Git repo, you can push your Heroku app onto their server using the very similar commands. This is easiest/simplest way to deploy apps on Heroku. GitHub Integration: If your repo is hosted on GitHub, then this is like a piece of cake, you can simply automate your whole deployment process. Dropbox Sync: If you're not hosting your repo for any reason and just saving them into your Dropbox folder, then also you're covered with the Heroku's easiest Deployment process. You can simply connect your Heroku ac...