Question about this Ai's identity (Solved)
Why is it when I question if this Ai model is 'Vicuna 13b', it responds "No, it's model GPT 3.1", vs when I question whether Vicuna is Vicuna 13b. (referring to the non unfiltered version, she's able to recognize her model type? I'm just super curious about how this model was trained to lose its identity.
Oh, nevermind folks. I've created the problem myself due to forcefully importing llama models using a specific batch file command which screwed up her import. Great job on Vicuna 13b unfiltered! Thanks for this release!
Btw, in case anyone is curious. Here's the batch command that I've used that caused a problem with my gpt 3.1 import of Vicuna.
These are the bad instructions:
call python server.py --model alpacino-13b-4bit-128g --model_type llama --share
Then resolved with these import instructions to launch my webui.
call python server.py --auto-devices --load-in-8bit --share --chat --groupsize 64
Just wanted to make a note of that to the community how this problem happened. <3 Thanks!