Medicine Custom LLM

Here we develop Custom LLM train on medicine data using LoRa techniques. I use Phi 1.5 microsoft LLM and it provide very good result. To learn about the custom LLM training and inference go through the given colab. I explain everything in detail in this colab notebook. What is LoRa PEFT (Parameter Effecinet Tuning).

Base LLM

microsoft/phi-1_5

Dataset

  MattBastar/Medicine_Details

Note: This dataset avialable at HuggingFace

Model Training Logs