Cognitive Computations drops DolphinVision-72b Our first multimodal model, and it's a banger! Built on llava-next and dolphin-2.9.2-qwen2-72b Created by qnguyen3 and Eric Hartford Massive thanks to Crusoe and TensorWave for the compute power! Shoutout to

We're bringing you zero-knowledge in a nutshell. In our latest blog, we cover the history of #zk technology and explore the intricacies of how the tech has evolved since 1989 -- when its first paper was published. Click the link to read more ⬇️ aztec.network/blog/does-zero…

Devnet is now live! 🎉 To celebrate, we’re launching Alpha Build, a series of three developer sprints with a USD $100,000 prize pool and the opportunity to deploy on the Aztec Network for the first time. Learn more and get involved ⤵️ aztec.network/blog/devnet-go…

Flux.1 Quantization Quality: BNB nf4 vs GGUF-Q8 vs FP16

Cognitive Computations presents Dolphin-2.9.4-gemma2-2b, curated and trained by Eric Hartford Much thanks to Google AI for the amazing base model, and Crusoe for sponsoring the compute! Dolphin is a general purpose chat model, uncensored, obeys the system prompt,

dolphin-2.9.4-gemma2-2b generates this faster than you can blink. ~2gb of ram

✨ New Model ✨ We’re excited to offer Dolphin 72B to Pro users. Venice’s most uncensored chat model yet, Dolphin’s datasets have been filtered to remove alignment and bias, making the model highly compliant with requests and system prompt instructions.

💡Model Spotlight 💡 Dolphin 72B is now available on Venice for Pro users Built on Alibaba’s Qwen2 72B, this model delivers uncensored and highly compliant responses Venice hosts models like Dolphin to provide users with unrestricted access to AI 🧵

The most uncensored chat model is live on Venice.ai (Mid weight, 72b parameter) h/t @cognitivecompai

Many investors still sidelined for DeFi, think they don't move as much/fast compare to other sectors like Memecoins or AI. When most DeFi protocols literally have refleixivtiy mechanism baked into the protocol. Quite a number of DeFi tokens have 5x from the Sep bottom with

1/ Anons: Let’s talk about why the current crypto AI agents (those bots claiming DeFi mastery) aren’t cut out for real financial moves on-chain. They rely on “RAG” logic and can’t handle complex DeFi tasks with any real reliability.

Uniswap v4 is here🦄 Users can LP on v4 through the Uniswap web app and swapping is rolling out over the coming days on web and wallet as liquidity migrates to v4 Live on Ethereum, Polygon, Arbitrum, OP Mainnet, Base, BNB Chain, Blast, World Chain, Avalanche, and Zora Network

Dolphin Mistral 24B Venice Edition is released Dolphin Mistral 24B Venice Edition is a collaborative project we undertook with Venice with the goal of creating the most uncensored version of Mistral 24B for use within the Venice ecosystem Dolphin Mistral 24B Venice Edition

Introducing Dolphin Mistral 24B Venice Edition 🐬🌅 Venice and Dolphin are co-releasing Dolphin Mistral 24B Venice Edition – the most uncensored AI model yet Let’s break down this release 🧵

Dolphin x Venice have released the most uncensored model ever Available to all users (free and pro) Try it without an account: venice.ai/chat

Ask anything. Privately. Our new Dolphin Mistral 24B Venice Edition model is the most uncensored model ever It refuses only 2.2% of user requests, beating out Grok, Claude and ChatGPT

The $27M Venice eco incentives fund is massive and nowhere near priced in imo Their first cohort of 50 devs/companies+ is bound to include some heavy hitters, Dolphin is already a world class team building vith venice. The number of DAU of venice has also been steadily

The Dolphin Telegram bot is now live Hosting Dolphin Mistral 24B Venice Edition for free to all Currently supports chat + coding configs & live URL scanning Usage guide below🐬🧵

Venice Uncensored just got better Together with Dolphin we’ve made significant improvements to Venice Uncensored, our most refined uncensored model available to all users Let’s break down what’s new 🧵

Dolphin Mistral 24B Venice Edition v1.1 is now live The model has been retrained on an updated dataset w/ reinforcement learning used afterwards to improve creative writing capabilities & further remove censorship based on feedback from the previous version Developed in