Flayute

Un psicólogo muy interesado en inferencia local
Anarco-telépata

[Flayute @flayute_]

AUDIO ON!
This is hypnotic...
@ItsmeAjayKV I loved this test!

[Flayute @flayute_]

I took your prompt and asked 27b to add ripple reaction to the audio in the pc...
It's very satisfying, thinking on adding some very light colour to it so it can discriminate low medium and high sounds...

[Flayute @flayute_]

Loving qwen 2.1 image @Alibaba_Qwen !!!
This is awesome. Running this quality in an rx7900gre 16gb is crazy
This is q5

[PrismML @PrismML]

Today, we’re announcing Ternary Bonsai 2 27B.

Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance.

Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use.

Ternary Bonsai 2 27B is available today under Apache 2.0.

[Flayute @flayute_]

@sudoingX 7/8 check
Just gotta know what to do with all this...

[Flayute @flayute_]

@remerikos @Alviseperez 7 vidas tiene un perro?
Alvise cabron!

[Flayute @flayute_]

@LottoLabs As someone already said this is weeks for token 🤣

[Flayute @flayute_]

Asking hermes to update llama cpp in termux so I can try k2 horizon 0.9b in my phone

[Flayute @flayute_]

@ymuchosmemes Yo colocando las bolas de billar en el triángulo...

[Flayute @flayute_]

@sudoingX 4gb lenovo p11 tab

[Flayute @flayute_]

@ashebytes Great music btw

[Flayute @flayute_]

@ashebytes Today I've been planning a human atlas for sedentary individuals with qwen 3.8 27b local.
This is just awesome to work with!
Thanks for sharing

[Flayute @flayute_]

@KyleHessling1 The difference with the base model in the thinking process is huge
This one just goes straightforward. This makes the task much faster and it also delivers ~5tks +. Over 50tks
Running iq4 xs m in a dedicated rx7900xt 20gb with 80k context
19gb overall

[Flayute @flayute_]

@KyleHessling1 Let's go! I'll try to fit the q4 in 20gb
With vanilla 27b I can fit 70k+ ctx enough for hermes
If it's more efficient in thinking, this one's gonna fly

[Flayute @flayute_]

@MiaAI_lab @QwenDevs Blink once for yes

[Flayute @flayute_]

@Hikari_07_jp There is real value, for ppl like me, and mostly like you, that had no idea in this matter and now you rock!
Owning your hardware and the intelligence it can bear, is far more interesting than using frontier intelligence

[Flayute @flayute_]

@QwenDevs The one I can run

[AJ @ItsmeAjayKV]

@Oluwaphilemon1 This is my post!
https://t.co/s0TzhQs76R

[Flayute @flayute_]

@Oluwaphilemon1 This is his post
https://t.co/VyGvsa9XTo

[AJ @ItsmeAjayKV]

@Oluwaphilemon1 This is my post!
https://t.co/s0TzhQs76R

[Flayute @flayute_]

@ItsmeAjayKV You deserve it!
Well done

[Flayute @flayute_]

@CuartoDelantero Li falta un tobogan

[Flayute @flayute_]

Is gemma 5 coming anytime soon?