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", ...

Flutter framework is the only framework that has come closer to satisfying this adage: "Write once, Deploy anywhere" Flutter apps could be literally deployed on all the major platforms that any business wants to that includes Web, Mobile(iOS/AOS), Desktop (masOS, Windows, Linux), which covers all the major platforms and there's really no need of writing platform specific code as Flutter framework does it under the hood for you plus there are a variety of plugins written by a thriving community that makes it even easier to develop apps. basically, with Flutter you have everything you need to develop apps without worrying about maintaining multiple platform codebases which is the main reason[read: lower costs] why companies opt for cross-platform frameworks in the first place.  listing out the pros&cons on a very high level: Pros: good documentation apps are fast since for ex mobile apps are not reliant on the Javascript bridge for communicating with the native layer op...