Covalent on LinkedIn: We're excited for the AI Engineer World's Fair on June 25-27. If you are… (2024)

🦙 Serving LLMs through API endpoints takes just a few lines of Python with Covalent. The platform hosts and orchestrates the deployment of accelerated compute for speedy GPU performance. 🚀👉 See this tutorial for serving up a Llama-based model to an interactive chatbot. Full Python code, leveraging Llama3 models from Hugging Face, Pytorch, and Streamlit, here: https://lnkd.in/e58UAqjE#ai #genai #inference #llm #llama3 Hugging Face #pytorch #huggingface Streamlit

18

See how easy it is to deploy an AI image generation service on Covalent Cloud. Serve up high-quality, photorealistic images into marketing and other application workflows, and leverage serverless GPUs for fast, scalable output.👉 See this tutorial for the Python code, using a Stable Diffusion based model RealVisXL from Hugging Face, running an Nvidia A100 GPU here: https://lnkd.in/gbXun7i4#ai #genai #inference #StabilityAI #SDXL #StableDiffusion Stability AI Hugging Face #pytorch #AIinAdvertising #CreativeAI #huggingface

22

Another great collaboration with Xanadu. See how easy it is to run GPU-accelerated quantum circuit simulations at scale with PennyLane and Covalent.

10

Deploying LLMs on Covalent Cloud is a breeze, so for this example, we're doing something a little different, and serving up Phi-3 and Mistral agents against one another in a war of words. ⚔️One AI agent is a tough customer, and the other AI agent is a sales rep for... AI-powered software, of course! 💬👉 See this tutorial for simulating a sales interaction using different LLMs conversing with each other. Full Python code, leveraging Phi-3-mini and Mistral-7B-instruct from Unsloth with vLLM, here: https://lnkd.in/ekz4RNB9#ai #genai #inference #llm #microsoft #Phi3 #vllm Unsloth AI Mistral AI

15

Tonight's can't-miss event in Toronto is theAI TinkerersMeetup. Our very ownAra Ghukasyanwill be speaking about fine-tuning and deploying LLMs on state-of-the-art cloud hardware using nothing but Python.Details:https://lnkd.in/gXsX_--B#AI#GenerativeAI#LLM#TechEvent

18

Building multi-agent AI systems?🚀 Covalent enables developers to create and easily manage multi-agent AI applications with complex agentic workflows. Dependencies are expressed as a seamless Python workflow, and each agent can be fine-tuned and deployed on their own hardware and environment based on model size and usage.👉 See this tutorial for developing a prompt refining application that uses multiple agent services dedicated to improve a specific aspect of the prompt, as well as distribute and consolidate work to/from the agents. Full Python code, leveraging Llama3 models from Unsloth and vLLM, here: https://lnkd.in/e-b4r6c8#ai #genai #inference #llm #llama3 #vllm Unsloth AI

19

Today we’re excited to announce Function Serve, Covalent's new inference service! 🚀Function Serve makes it extremely simple to productionalize any AI/LLM or acceleratedcompute application with only a few additional lines of Python. Covalent handles all the underlying infrastructure for you.With Function Serve, you can:- Train, fine-tune, and serve any AI model in a single Python notebook- Create and easily manage multi-agent AI applications with complex interactions- Productionalize and scale any application or workflow that leverages accelerated computing (simulations, digital twins, scientific computing, etc.)👉 Read our blog "A simpler way to develop any AI application imaginable" https://lnkd.in/gypp2TDU🤖 Build a sophisticated multi-agent AI application entirely in Covalent, leveraging the workflow management capabilities alongside its serverless infrastructure: https://lnkd.in/ghsdqKQG💻 Deploy a Llama3 model for inference: https://lnkd.in/gGktPbiQ💎 Fine-tune LLMs w/ Covalent:https://lnkd.in/gaJiaFyt🌀 Realistic image generation using Stable Diffusionhttps://lnkd.in/gbXun7i4Sign up to Covalent today and get up to 30 free hours of GPU compute: https://lnkd.in/etsadYX6

49

1 Comment

Attending VivaTech this week? Don't miss our presentation on Wednesday, May 22 at 9:50-10:20am, and be sure to meet us at the Canadian Pavilion S21.#Canada #ScaleAI #vivatech

4

Attending PyCon US? Learn how to build an AI foundry with autonomous agents using Covalent Cloud, and deploy your next killer app with just a few lines of code.Don't miss "No Data? No Problem: Zero-Data Model Training with Foundational Models" presented by Santosh Kumar Radha and Ara Ghukasyan at 11:30am today.👉 Github repo: https://lnkd.in/essje6nvVisit booth 221 to pick up some swag!Python Software Foundation #PyConUS #PyCon #python #llm #genai

26