AI should be accessible to all, GPU-poor or rich. This repo contains well explained code to finetune Smaller LLMs such as Qwen1.5-0.5B and Mistral 7B in a free Google Colab Notebook.
All of this can be replicated in a Google Colab free instabnce with T4 GPU.
Contributions to this repository are welcome. If you find any issues or have suggestions for improvements, please open an issue or submit a pull request.
This is Apache 2.0 license.