Skip to content

phenogpt-llama2

Latest
Compare
Choose a tag to compare
@quannguyenminh103 quannguyenminh103 released this 02 Apr 20:24
· 23 commits to main since this release
0116e91

This is PhenoGPT model fine-tuned on Biolark dataset, based on falcon-40b, llama1-7b, and llama2-7b base models. In the phenogpt.zip file, you will find three subfolders, one for falcon model, one for llama 1, and the other for llama 2 (default). These three folders include (lora) adpaters' parameters and configurations. You can directly reload the fine-tuned model from this directory. Please refer to the notebook for details.