With which quantization ? I have a 16GB GPU, running Qwen3.5 9b UD_Q8_K_XL basically max out the VRAM usage. Maybe a MOE model fits you. Gemma 4 e4b is one with a total 8b.

And what is the token speed you are getting ?

Thanks

Where is the screenshot from ?

I guess you could wait until it is cracked. But by then, I think it is best to use a old pc.

You say also if RAM can be upgraded. Consoles are not like PCs in terms of customizations. Only the old ones that have been cracked. You can’t.

But your enthusiasm is there. That’s cool.

Good old games They sell them DRM free.

I think you meant librewolf ? And if so, I believe it is desktop only.

Yesterday I needed this. Will install this. Thanks.

May I ask: have you noticed if the prompt processing speeds shown in llama-bench are vastly different from llama-server ? I have hundreds of tokens of difference.

You mean Gemma 4 ? You read in his discord ?

I confirm the same and it works now. I set it to maximum because fewer reasoning effort tokens cuts it directly.

Thanks

I know it, seeing it in models titles.

Is uncensoring oneself a LLM difficult ?

Did you try any ? Because, I tried iglors and mradermacher, I got refusal to make a pipe bomb. Their answer are funny because they say to study academic engineering instead, lol. Still a refusal. I will try this one.

The Arch Linux forum make people do that.

Someone may have the same question in the future and there will be answers. You not responding is not that bad but it is even better that you do and provide an update to your situation, if you wish.

No. In fact, that is nice. I should try.

Where did you implement ? How did you add rate limits ? And you said “almost all”, who are those you can’t block ? Of course, you share if you want to. I don’t mind. Thanks.