Cartyisme

@sudoingX tok/s after a few months has become my benchmark, 20tk/s might as well have my agent chiseling in stone.

you have a GPU. you want to run a local model. what's the first thing you search for?

@TheAhmadOsman tbf I was having issues with tool calls, as it seems they were being wrapped w/ vLLM. And llama.cpp solved the issues. 5090 user here.

DGX Spark uses unified memory > 273 GB/s RTX PRO 6000 delivers > 1.8 TB/s (1792 GB/s) If someone told you they’re comparable, they’re wrong And this is exactly why llama.cpp isn’t the right tool here Try vLLM or SGLang on a GPU and you’ll see very different results

@stevibe Qwen3.5 turned it up a notch for sure, with a lot of exciting models coming down the pipe. Very exciting.

@cartyisme 100%. Luckily we still have time to fill the gap and learn how to set them up. Open-source progress is wild right now!

@z0age @NadavAHollander if u n nadav are cookin, i want in

so @NadavAHollander and I have been cooking on a new project if you’re regularly trawling your feed on the hunt for fresh information but are drowning in a firehose of unrelated slop… then you should reach out! we still have room for a few more curious minds in the first wave

@sudoingX @mamboussa this guy is awesome, love the insight, having gone through all of this myself, ollama is not the way. even gave u q4 quantization command for ur kvcache. tho 132k context seems ambitious, but im not running a 3090.... try Q3_K_M from @UnslothAI good trade off here

llama.cpp is the way. grab the Qwen3.5-9B Q4_K_M.gguf from huggingface, compile llama.cpp with CUDA, and launch with: ./llama-server -m model.gguf -ngl 99 -c 131072 -np 1 -fa on --cache-type-k q4_0 --cache-type-v q4_0 --host 0.0.0.0 then install hermes agent and point it at localhost:8080. dm me if you get stuck.

Hi @sudoingX im going to install Agent Hermes on my desktop with a 3060 12gb, can you help me to configure it with Llama.cpp or Ollama ?

@zjbrenner im literally watching this right now lol

Spending my Friday night watching Dinosaurs. You?

@peekcell i just meant to say, most AI pixel art or pixel art made in illustrator will break when scaled to nearest neighbor, making it clearly not "real" pixel art

@cartyisme Haha this works if you want to preserve the details with up and downscaling the pixels yeah!

As Ordinals support sunsets elsewhere, FAUV•TOSHI feels extremely prescient... tradeable on @OpenSea, inscribed on @Bitcoin, minted on @base. opensea.io/collection/fau…

@cartyisme @opensea

@javpixel_art @opensea updated

gm✨ Hi @opensea , you guys are showing the same collections since 1 month on Base Homepage. The module says "This Week curated collections" , however, the same collections are there week after week, some of them even without real volume in months! Is someone even paying attention?

Getting your opensea verified was the biggest flex back then. Shoutout @cartyisme for all the help over the years x.com/coinbilly/stat…

@AutismCapital Slapstick, or Lonesome No More!

We've said this before but it bears repeating. Establish your tribes NOW. The world is going to become INCREASINGLY tribal moving forward. It will also become exceedingly violent as systems fail. The world will unfortunately adopt an us vs them mentality. Establish your tribes.

The world is unrecognizable these days. Humans have become different. Everyone is being driven by different spirits than they previously were. It's only going to become even more unrecognizable. The idea of what it means to be a human, a citizen, an identity, is being challenged.

hoping to do the same, but cant find the power cord D:

@Maniinng @opensea Fixed it :)

@cartyisme @opensea opensea.io/item/ethereum/…💚

A few days have passed since this fake collection showed up. I shared it in the @opensea server and nothing, asked people to report it, and still nothing. How do you even get this kind of stuff taken down 🥲

@Maniinng @opensea Can you link me?

Thank you so much!!! Its being hard to contact, I have an open thread about a token that was sold, but on OpenSea it’s still listed and I still appear as the owner. I know you’re extremely busy, but it was amazing back when there was a support channel and things got fixed as we reported bugs.

@AspynPalatnick @netprotocolapp geeks > mops > sociopaths 2016 > 2021 > 2025 👀

A group chat I’m in was talking about how NFTs and memecoins are no longer punk Makes me think @netprotocolapp may be the most punk thing in crypto today Apps with centralized servers/DBs? Mainstream Censorship-resistant decentralized apps with all data stored onchain? Punk

@badte_eth @SteveKBark @HollanderAdam Mind shooting me a DM? Happy to try and help get this figured out and troubleshoot w/ you.

gm @SteveKBark @HollanderAdam I'm having persistent issues updating the collection overview images for opensea.io/collection/ska… If you could point me in the right direction to get it resolved I would greatly appreciate it