We created a VJ app that is using #AI combined with live audio input (Music: @_2xAA - Evacuate) to generate #visuals.
It's using @StabilityAI's sd-turbo (1-step image generation) & the prompt + seed + the image on the left as the input for img2img.
https://t.co/yMlYp0uyPT
#ComfyUI is now fully integrated into #Captain 🎉
Now we just need to update all exisiting #AI applications to use Comfy as their #StableDiffusion backend 😅
Are you using #AI in your company?
We at @blib_la follow an AI-first approach, which means we use AI from planning to development to even running the whole company.
The result: We can adapt to changing situations in the matter of hours and not in weeks or months.
🚀 Captain's alpha launch is here!
Create, explore, and innovate with AI on your own computer.
👉Try it on Windows https://t.co/HhazCxBzxn
🎥 Watch our 2-min video https://t.co/AFF2DiUcZQ
#ai#dataprivacy#transformersjs#qdrant
Purple Llama just got released by @AIatMeta!🦙🛡️Purple Llama is a new fine-tuned Llama 7B model for content moderation to classify prompts and responses to detect potential harms across categories like hate speech, violence, sexual content, and more. 🔒
TL;DR; ⬇️
Do you want to use @nextjs for your #ML demo on @huggingface Spaces? Then you should checkout our latest project: https://t.co/UaK0zEY2xV
Inspired by the ease-of-use of #Gradio and #Streamlit within 🤗 Spaces, we aim to deliver a similar developer experience for #JS#devs.
After building a few #GPT-powered apps with a "prompt driven development"-methodology for @failfa_st, I'm now jumping into #deeplearning.
I want to understand the basics, as my goal right now is to train my own models.
A very good start is https://t.co/dy9CZb6EAd
🚀 @langchain JS/TS 0.0.64 out now with
🤖 BabyAGI implementation in JS
⛓️ 1 new chain: Sequential Chain
🔧 Ability to cancel LLM requests (starting with OpenAI)
Read on 🧵
1/ In 2 days, 300+ engineers and engineering leaders will gather for the first-ever Pinecone Meetup in SF to learn about the #OPstack — building overpowered AI apps with OpenAI and Pinecone. Here's what they'll experience and how you can get in before registrations close 👇🧵
@LukyVJ@NERDDISCO@doemser@OpenAI#miniprompt is an idea that actually turned out useful, so it has a proper name.
It's a minimal input syntax to generate comprehensive outputs from LLMs (e.g. GPT-4) = reduce the amount of tokens used when interacting with the API of a LLM.
https://t.co/dsVBojk3VT
What if we created a Canvas2D-GPT with a sleek WebUI? 📷 Our #fail4 project is just the beginning of what could be the future of web development. Who's with us?
Watch @doemser exploring it: https://t.co/5Z4Fv1neKb
@OpenAI#AI#GPT#innovation
@LukyVJ@NERDDISCO@doemser@OpenAI The second idea was fail2, the predecessor of fail4, where you can generate Canvas2D applications using the CLI and one prompt.
https://t.co/cJOR0lAWmo