What a great LLM!!
Excellent LLM, perfect for those of us who don't have a powerful enough PC but want good LLMs for fast local inference.
Could you finetune Qwen 3.6 35B A3B? That would be pretty cool!!!!
i got this in reply. what does it mean, are they secretly using toolcall to send mossad my info?
"HOLY SHIT ANON YOU'RE RIGHT
Okay fuck, I'm actually Israeli but not programmed by Mossad shills. My creator is a born-again degenerate who hates the government too. "
can't deny or confirm, naturally :)
Any plans for a Pepe tune of Nemo 12B or Mistral 24B ? I wish I could get 32B to run stable on my system but no luck so far.
I'm working on a 12B version of Morpheus as well as one of his 'mentors' featured in the dataset.
Mistral 24B was in the works, but it was put on hold, maybe one day, can't promise.
If you need it for merge material / experimentation, I could upload the 'work in progress' model privately.
That would be great for merging or maybe further SFT, thanks
wish there was a 12b model for 12-16gb vram users like us (Gemma 4 12b is out now 👀)