General discussion.
For potential quant-file issues and related.
Very good. Reasoning now really helps with logic. I had a character to whom I gave some items in a backpack at the beginning of the story. Among the items was a magnifying glass. In the middle of the story, I came to the forest and asked the character to make a fire. Mag-Mell-12B-R1 did not always choose the obvious item, even when I hinted to the character that today was sunny weather. Violet_Magcap-12B almost always chose the magnifying glass among many things.
Yeah, the community response has been pretty solid overall on this one. Glad to hear your experience has been positive! I’m hoping to get a v2 out sometime next week, with a bit more focus on RP-specific reasoning training. The reasoning format might change a bit— using just think tags without answer tags helps save a good chunk of tokens. I’m also considering shifting the reasoning trigger over to a system prompt instead of relying on user prefixes, just to keep things more controllable overall. With a bit of data manipulation, I could even tie in control over reasoning length as well.
I’m also considering shifting the reasoning trigger over to a system prompt instead of relying on user prefixes, just to keep things more controllable overall.
That would be nice.
So I´ve been getting some amazing generations with this model so far, however I´m pretty sure I am doing something wrong because think and isntruction text is bleeding into the prompt.
I downloaded the json files but to be honest I have NO idea where I should import them.
Can you point me properly?
Thx
@blackroseyagami
https://files.catbox.moe/zdwrsr.json Latest Master Import
https://files.catbox.moe/ssupgx.json Reasoning Quick Reply Preset
@blackroseyagami
https://files.catbox.moe/zdwrsr.json Latest Master Import
https://files.catbox.moe/ssupgx.json Reasoning Quick Reply Preset
Welp, this has been amazing so far. Even with some character cards that have been very half assedly made by me it returns very good replies and generations. I have been looking for quite some time for a model I can locally run with my 4060 without having to overtweak every setting.
EDIT: I have a character card that I´m pretty sure it just very badly created by me and it tends to repeat text A LOT, it takes the last 1 or 2 paragraphs and reapeats them at least 4 more times, sometimes more if the token count permits it.
Interesting...
At some point, still haven't figured out what triggers it. It starts a resoning prompt and sorta resets everything you have so far during the roleplay and starts as if it were a new chat. I´ve had this repeat itself on different characters. I´ve modified the text and OCC comments to see if it retakes where the last message was left off but it just doesn't budge. Maybe a hardware limitation on my part?