this model is... wow. Just wow.
This is by far the best 13B I ever had the pleasure of testing. Using the same prompt and ooba, it feels far beyond even the predecessing uncensored wizard models (gptq, exllama or ggml) with excellent speed as well. Is this because of the new format or is Llama2 such a large improvement? In any case, fantastic work! Have been a big fan of your uploads since (almost) day one!
Llama 2 is a very big improvement and wizardlm has been trained on new, better data as well! So pretty cool.
How do you use this model using the tools from h2ogpt? I cannot even load the model
How do you use this model using the tools from h2ogpt? I cannot even load the model
use LL Studio to use this, I am a novice and have zero experience with all this technology and i was easily able to use it. all you do is download and use it. there are a lot of videos on how to do it, search for this exact model and download.
How do you use this model using the tools from h2ogpt? I cannot even load the model
use LL Studio to use this, I am a novice and have zero experience with all this technology and i was easily able to use it. all you do is download and use it. there are a lot of videos on how to do it, search for this exact model and download.
Thank you for the tip. I'm new to this as well and it seems a bit daunting. So using LL Studio is the best way?
How do you use this model using the tools from h2ogpt? I cannot even load the model
use LL Studio to use this, I am a novice and have zero experience with all this technology and i was easily able to use it. all you do is download and use it. there are a lot of videos on how to do it, search for this exact model and download.
Thank you for the tip. I'm new to this as well and it seems a bit daunting. So using LL Studio is the best way?
just seeing your reply, Yes LM Studio is by far the easiest way to use it, all you do is download the software and the download this model or any other model to be used within it. IT's pretty straight forward