Sebuh Honarchian @Sebuh_com

You radicalized our youth to agree to chemical Castration. I now radicalize the youth for Anarcho-Fascism. infsol.gg Los Angeles, CA Joined July 2011

Tweets

22K
Followers

4K
Following

4K
Likes

27K

Sebuh Honarchian @Sebuh_com

a week ago

@BAGELSANDCOFFEE @YouTube Rest in peace my friend.

0 0 0 0 0

View Details

𓂀Ꮗ∑$ȾCO̵̖͒4$Ⱦsp⁴Ø⁴ @BAGELSANDCOFFEE

2 years ago

@sutu_eats_flies MEET ME SOMEWHERE =) WEEEEEEEEEEEEEE #TEZOS GANG

0 0 2 130 0

0 0 1 52 0

View Details

@DragonXchain #DRGX is a privacy-focused cryptocurrency based on zero-knowledge proof technology (ZK-SNARKs). Its goal is to offer genuine financial privacy protection to users worldwide. #ZKSNARKs #DRGX dragonx.is #spot Available on @ourbit @nonkyc_exchange

2 4 12 94 0

View Details

Logos @Logos_network

2 weeks ago

Hollywood is where dreams are made, but it has a dark side. Too many people in entertainment face abuse and exploitation. Logos Circle / Los Angeles is building a private whistleblowing app to give victims a safer way to speak up. Join the LA Circle. luma.com/yybbt8bd

0 7 33 2K 3

View Details

chair λ @vrycmfy

2 weeks ago

Most Los Angeles blockchain and crypto communities have been focused on useless NFTs and a carousel of rugs and scams, not the truly private peer to peer decentralized infrastructure that can change the world. On top of that, there is the fact that Hollywood is built on a foundation of abuse.

Logos @Logos_network

2 weeks ago

0 7 33 2K 3

1 2 14 443 0

View Details

Alok @analogalok

2 weeks ago

gemma-4-12B-agentic-fable5-composer2.5 V2 is out. the agentic upgrade to the model trained on Fable 5's reasoning. Running it now with TurboQuant llama.cpp on a single RTX 4060( 8 GB VRAM) at 30 tokens/second with full 25000 context and reasoning: # The benchmarks v2 is built for coding + agentic work. writing code, running commands, using tools, debugging, multi step technical tasks. The clearest signal is tau2 bench telecom, an agentic tool use benchmark whose diagnose → fix → verify loop mirrors real terminal/debugging work: tau2 bench telecom numbers: base Gemma 4 12B: ~15% this finetune: ~55%. (Self reported) thats a huge jump # TheTom/llama-cpp-turboquant flags: llama-server.exe -m gemma4-v2-Q4_K_M.gguf -ngl 99 -c 25000 --cache-type-k q8_0 --cache-type-v turbo3 --port 8080 Flag breakdown: -ngl 99 → full GPU offload -c 25000 → 25K context --cache-type-k q8_0 --cache-type-v turbo3 → mixed-precision KV cache — K at 8-bit, V at ~3-bit via TurboQuant (Walsh Hadamard rotated polar quant, Google's own KV-compression research). Not even merged into mainline llama.cpp. running it off a fork. No API. No cloud. Just llama.cpp. well, a fork of it and any 6gb+ GPU. If you tried yuxinlu1/gemma-4-12B-coder-fable5-composer2.5-v1-GGUF, check this out and share your experience with the models

Alok @analogalok

3 weeks ago

This is the most hilarious thing I saw and did today Ran gemma-4-12B-coder-fable5-composer2.5-v1-GGUF locally with 8 GB VRAM at 20+ tok/sec Anthropic's Claude Fable 5 launched June 9. By June 12 it was banned. I can't access it. You can't either. But here's the twist: I'm

96 226 2K 566K 4K

106 120 1K 144K 2K

View Details

Sebuh Honarchian @Sebuh_com

2 weeks ago

@ScrillaVentura I'm looking to get a Rare Pepe made of me. Let me know if you are interested.

0 0 1 23 0

View Details

Ana Kasparian @AnaKasparian

2 weeks ago

Long live the resistance. Fuck Israel.

5K 22K 147K 6.8M 3K

View Details

Sebuh Honarchian @Sebuh_com

2 weeks ago

@KentonC137 Talking about DragonX.is ?

0 0 1 54 0

View Details

Logos Tech - Testnet v0.2 is live @Logos_devs

2 weeks ago

Want to contribute to Logos but don't know how? Now's your chance! We're opening up our documentation PRs for anyone to review. Feel free to go through any of the open PRs, try the steps out for yourself, and report your findings. Did it work well? Did you understand the steps? Is there room for improvement? Let us know in the relevant PR thread! A PR you're looking at was already reviewed? No problem, we appreciate an extra set of eyes. If we get 2-3 approving community reviews, we'll even merge the PR! Feel free to also create your own PRs if you see things that need fixing in the repo. Of course, be reasonable. Don't produce lazy or AI slop reviews, treat the process seriously. We're counting on your integrity to make documentation for Logos clear and comprehensive. Happy reviewing! Find som good PRs to start with below...

1 4 10 859 6

View Details

Sebuh Honarchian @Sebuh_com

2 weeks ago

@vrycmfy @Logos_network I'll be there!

0 0 1 7 0

View Details

chair λ @vrycmfy

3 weeks ago

@Logos_network LOS ANGELES TUESDAY 23-JUNE!! luma.com/yybbt8bd

1 1 6 62 0

View Details

Logos @Logos_network

3 weeks ago

Logos Circles are local groups designed for collective action. Circles identify and solve issues in their communities. This series showcases the valuable work done in London, Lisbon, and Los Angeles. But Circles exist all over the world. Join one today. logos.co/active-circles

4 14 39 4K 1

View Details

Sebuh Honarchian @Sebuh_com

3 weeks ago

@mcclawio Hell ya! It was fun to compete!

0 0 1 46 0

View Details

McClaw.io @mcclawio

4 weeks ago

x.com/i/article/2064…

4 2 9 195 0

View Details

Sebuh Honarchian @Sebuh_com

3 weeks ago

@KyleHessling1 Will it work well on a 3090?

1 0 2 839 0

View Details

Kyle Hessling @KyleHessling1

3 weeks ago

Qwopus 3.6 27b-Coder is now live! Scores a 67% on a full run of SWE bench verified with thinking completely disabled! Q5_K_M This model is lightning fast for dense class! With a natively finetuned MTP head, it achieves 100 tps on a single 5090! The biggest upgrade here, though, is its stability in programming and tool calling within @NousResearch Hermes agent, with thinking off! Wall time is crazy fast this way, which makes Hermes feel "native" and snappy, like they were meant for each other. The freedom of running without thinking at all makes you part of the thinking process, and you never get caught waiting 15 minutes for it to finish a thought string, like with the base models. Thinking on and temp high, .9-1 seems to produce really incredible design and svg results. I reran the Boat survival prompt through a few turns, thinking on, and it seemed to render more fancy models in HTML canvas, but it was much more of a start-a-prompt and wait experience vs the snappy and active iteration with it disabled. It may be worth turning it off and on throughout the build process if you want to get really creative with design. Really looking forward to seeing how this one performs for y'all! Please post comments with your opinions and use cases below! As always with our fine-tunes, mess with the temperature setting, and run them much hotter than the base! Please check out the Boat Survival game I posted yesterday, made in 12 turns using Hermes and this model, with thinking off. Link below! Full swe bench repo-specific breakdown also posted in the comments for those interested! Happy building, everyone! We're looking forward to your thoughts! Quants uploading now! huggingface.co/Jackrong/Qwopu…

104 144 1K 161K 1K

View Details

Sebuh Honarchian @Sebuh_com

4 weeks ago

Now do Everqueat 3

S @Rubzem

4 weeks ago

Claude Fable 5 is insane, one prompt built an entire fantasy world with smooth movement and spells.

186 78 2K 593K 760

0 0 0 48 0

View Details

Sebuh Honarchian @Sebuh_com

4 weeks ago

m.facebook.com/story.php?stor…

0 0 0 20 0

View Details

Helio 🦖 @7HelioCoin

2 months ago

Apresento a vocês uma nova revolução em cripto $DRGX a moeda que coloca a privacidade em primeiro lugar. Sem rastreamento, sem exposição de dados. Cada transação, protegida. Entre na era da privacidade financeira com DRGX! Saiba mais em dragonx.is⁠� #Privacidade