LLM Finetuning Hub is a platform that provides code and insights for finetuning large language models (LLMs) to suit different use cases. The Evaluation Framework consists of four pillars that help evaluate LLMs holistically. The platform offers ready-to-use scripts for tasks such as finetuning LLMs on proprietary datasets, hyperparameter optimization, and more. The process of fine-tuning LLMs is made easy with step-by-step instructions. The roadmap of LLMs covered includes Flan-T5, Falcon, RedPajama, Llama-2, OpenLlama, SalesForce XGen, Mosaic MPT, Cerebras, Writer Palmyra, OpenAI GPT-3.5, Google PaLM, and Inflection Pi. Contributions to the project can be made through the Git workflow. For inquiries, there are contact details provided.
https://github.com/georgian-io/LLM-Finetuning-Hub