Smaller version?
You're cooking. Can we have smaller version maybe gemma 3 12B?
Greetings. In version V4, I didn't plan to release a smaller version, at least not anytime soon. I really want to wait for Qwen3 and see what it's capable of. If it turns out to be generally worthy, I’ll definitely train a smaller model in V4.1. However, if it performs poorly, then after some time, I’ll train Gemma3 12B instead. I’m sorry, but my resources are limited, so I have to be selective about training models. I truly appreciate your understanding.
Any update re: the next iteration of this model series? Really like V4, looking forward to whatever you've got planned.
Any update re: the next iteration of this model series? Really like V4, looking forward to whatever you've got planned.
Greetings!
I am currently on a bit of a vacation. I've been working on the CardProjector series every day for six months straight, from V1 all the way to V4, so I decided to take a short break to recharge my batteries before tackling CardProjector V5. I plan to experiment with Qwen3 because it looks very promising at the moment. Stay tuned for updates!