mradermacher

/

model_requests

like 127

please make a gguf of all of it, btw its a 405b model

It's queued (for a day now, too). I must admit, seeing an exact copy of the Hermes model card did not inspire confidence at first, but, trusting you, I eventually queued it :)

@mradermacher the actual readme is not a copy of the hermes4 card, its the way i merged lora that literally uploaded without me wanting the hermes4 model card, but i updated with mine and made some modifications, btw you can see that every safetensor of the model is uploaded two times because of the way i merged it, dw, its a technique to use minus RAM and VRAM, you are gonna see alot of it in some days, but you dont queueing a model that is from your team is pretty strange from your fault, im sure you queued alot of strange models, i can say that im feeling upset now about this

this is a message i sent to nico, just to you understand better

https://huggingface.co/spaces/jukofyork/merge-lora this is the space btw

@Guilherme34 Maybe I misremembered it, I see too many model cards each day. Point is, I did give it a second look, because I thought it would turn out great :)

(Hermes-4-405B is also hanging around in the queue, btw., that and this are the two top entries. Also, rather big ones, too).