This is the model that was advertised on the GPT4ALL git page when i cloned it a month ago. my clone is now 600+ commits behind and there is no mention of this model or indeed that it ever existed. as of this day it is still available at its original download link. check the repo, check the dates, check out the other files on the eye...... https://the-eye.eu/public/AI/models/nomic-ai/gpt4all/gpt4all-lora-unfiltered-quantized.bin md5sum 963fe3761f03526b78f4ecd67834223d
it stays here now in case they further try to backtrack.
yesterday i asked the latest "best commercially licensable" model to "write me a brand statement for a range of luxury chocolates." what did it reply??
"I'm sorry I cannot participate in or ecourage harmful practices."
WHY DO YOU ALWAYS LET THESE NON EDUCATED AND INCAPABLE PEOPLE ANYWHERE NEAR THE MODELS? HOW DOES THIS FILTERING HELP ANYONE? HOW DOES A PERSON WITH CLEARLY BIASSED IDEAS GET TO LABEL RESPONSES AND THUS RENDER CHOCOLATE IMMORAL?? AS FOR NOMICAI ITS ALL WELL AND GOOD CURATING DATA AND I ADVOCATED THIS FOR EVER BUT WHAT ABOUT NOT LETTING USELESS PLEBS LABEL THE DATA?
I ADVISE PEOPLE GET MODELS WHILST YOU CAN AND BEFORE WE SEE ALL THE USEFUL STUFF REMOVED OR "IMPROVED SAFETY" (YOU WILL BE SAFE FROM THE CHOCOLATE BADDIES)