Great minds think alike
I did a mythomax-chronos-xwin based merge a little bit ago. Yours looks better ofc :P If you have time a 2k quant would be awesome? ( as until we get 13b mistral my mobile dgpu is arse for 13b models)
Only if it's no trouble, as I bet thebloke will probably get to this one sometime :)
Legend :)
@Undi95 @BlueNipples , I did leave a message in another model comment section too, yet hehehe... as I read the comments here again about Xwin, I couldn't help but mention it here too.
I was wondering if you guys have tried this model: https://huggingface.co/TheBloke/Llama-2-70B-Ensemble-v5-GGUF/tree/main, Ensemble and what do you think of it?
@Undi95 @BlueNipples , I did leave a message in another model comment section too, yet hehehe... as I read the comments here again about Xwin, I couldn't help but mention it here too.
I was wondering if you guys have tried this model: https://huggingface.co/TheBloke/Llama-2-70B-Ensemble-v5-GGUF/tree/main, Ensemble and what do you think of it?
I don't work a lot with 70B as it's a pain in the ass, but I didn't tried it nope, I will if I have the time. Thank you!
@Undi95 @BlueNipples , I did leave a message in another model comment section too, yet hehehe... as I read the comments here again about Xwin, I couldn't help but mention it here too.
I was wondering if you guys have tried this model: https://huggingface.co/TheBloke/Llama-2-70B-Ensemble-v5-GGUF/tree/main, Ensemble and what do you think of it?
I don't work a lot with 70B as it's a pain in the ass, but I didn't tried it nope, I will if I have the time. Thank you!
Btw, I just found out that the ensemble family of models were also released in 13B variants. Here are the links for them:
13B V5 Ensemble model: https://huggingface.co/TheBloke/Llama-2-13B-Ensemble-v5-GGUF
13B V6 Ensemble model: https://huggingface.co/TheBloke/Llama-2-13B-Ensemble-v6-GGUF
if that helps in any way π .