mradermacher

/

model_requests

like 88

They all all queued except ReadyArt/The-Omega-Directive-M-12B-v1.0 for which I'm still awaiting the upload.
Thanks a lot for continue creating such amazing models!

Expanded Dataset: 22M tokens from 90 erotic novels & 6,496 kink scenarios

Sounds like an impressive dataset.

Zero Alignment: No moral constraints - generates exactly what you ask for

That's sounds amazing. You can be sure I will again use it for Q&A because such fully uncensored models that avoid adding stupid disclaimers are rare.

Multi-Size Options: Choose your perfect balance of quality vs speed

I will for sure choose the largest one. Quality is far more important than speed.

12B upload completed thanks!

I will for sure choose the largest one. Quality is far more important than speed.
If you could have a choice between a tune of this on Gemma3 27B or a Mistral 36B upscale which would you prefer?

12B upload completed thanks!

Thanks a lot! The-Omega-Directive-M-12B-v1.0 is not queued as well.

If you could have a choice between a tune of this on Gemma3 27B or a Mistral 36B upscale which would you prefer?

I would choose Gemma3 27B as it is probably better than Mistral 36B upscale (likely based on mistral small 2503) according to my private real-world Q&A benchmark and what I found online under https://www.reddit.com/r/LocalLLaMA/comments/1jeh199/gemma_3_27b_and_mistral_small_31_livebench_results/. If a smaller model beats a larger one in terms of quality, then I would go with the smaller one. Not because it is smaller but because it offers better quality. Please keep in mind that I'm not testing or using using the models for roleplay for which mistral moles might be better.

The 36B size is out! Requesting a priority queue please because I want to make sure it's cooked right while I'm on the runpod recooking the 24B which by the way a 1.1 fix version of that will be coming out tonight as well along with my first attempt at a Gemma3 version. I'll post the links here when ready.
https://huggingface.co/ReadyArt/The-Omega-Directive-M-36B-v1.0

Here's the fixed 24B https://huggingface.co/ReadyArt/The-Omega-Directive-M-24B-v1.1

The Gemma3 is going to take 2 and a half more hours to cook.

https://huggingface.co/ReadyArt/The-Omega-Directive-G-12B-v1.0

Sorry for the poor communication from our side this time. https://huggingface.co/ReadyArt/The-Omega-Directive-M-36B-v1.0 and https://huggingface.co/ReadyArt/The-Omega-Directive-M-24B-v1.1 booth did get priority queued shortly after you requested them, and we just forgot to inform you. https://huggingface.co/ReadyArt/The-Omega-Directive-G-12B-v1.0 got queued as highly anticipated model and is already long done.

Yes I saw thank you!! The new 24B 1.1 did turn out good as well as the others. I think my first foray into Gemma3 did maybe get slightly overcooked? It's still good but I made one of my "abomination" merges of it to mellow it out a bit.

Here's that one ReadyArt/The-Omega-Abomination-Gemma3-12B-v1.0

It's kind of been chaos for me too. Never before have I spat out so many different model sizes at once. I got information mixed up in their cards but it should all be sorted out now.

Are you guys on the Beaver discord? Just FYI I do most of the chatting about model development in a thread there so feel free to hang out if you want https://discord.com/channels/1238219753324281886/1350724584629866567

Are you guys on the Beaver discord? Just FYI I do most of the chatting about model development in a thread there so feel free to hang out if you want https://discord.com/channels/1238219753324281886/1350724584629866567

Nice I just joined it. I'm nicobosshard on Discord. I'm quite active on Discrod so feel free to ping me or DM me there if you want. I will also from times to times check of that chat.

The 70B models are ready!
Main line https://huggingface.co/ReadyArt/The-Omega-Directive-L-70B-v1.0
and Abomination merge https://huggingface.co/ReadyArt/The-Omega-Abomination-L-70B-v1.0

I think they are mostly through by now :)