Lepton AI

OpenAI has rolled out native voice mode support for GPT4, but what about your own llama models? We are excited to release our new version of Lepton LLM engine, offering integrated voice mode input and output within one optimized inference call! 🧵

Navigating the complexities of the GPU market can be daunting. We understand the need for a clear and comprehensive guide to make informed decisions. To learn more about how we help our customers run H100: blog.lepton.ai/the-missing-g…

We're open-sourcing GPUd, an AI-native GPU management utility that reduces GPU cluster unavailability by 4x. Built by experts from Meta, Alibaba, and Uber, GPUd automates monitoring, diagnostics, and issue identification for GPUs—ideal for cloud or on-premise, at any scale. 🧵

Below is the same example and using (almost) the same code from the OpenAI blogpost and Llama3.1 got the correct result (and also the step by step thought process!)

Huge shout out to @FishAudio for the awesome voice generation model!

"The only reason to give a speech is to change the world." We are bring native, real-time voice generation to all open source LLMs. Instead of separate, duck-taped modules, we build one single engine to deliver both the text and audio with time to first audio around 300 ms. 🧵

1/5 We are super proud to announce FastGPU, our lightweight IaaS as your go-to solution for simple, reliable, cost-effective cloud GPUs. Try it out at fastgpu.com today! 🧵

2/5 As part of the FastGPU launch, for a limited time, we’ll be offering RTX 4090 GPUs at $0.65/hour. This is a GPU widely liked by many developers, running workloads like graphic rendering, numerical computation, web3 computation, and many others. We hope you enjoy it!

3/5 We showcase the capability of fastgpu by running the most recent Stable Diffusion 3 medium model - you can try it out completely free at fastgpu.com/sd3.

4/5 We are super proud to partner with @exa_bits to combine high-performance, cost-effective GPUs and a fully cloud-native platform together. It takes a multitude of expertise to make AI cloud a reality, and we are super excited about the partnership.

5/5 FastGPU’s software platform is built on Lepton AI’s technology. Lepton AI builds a modern AI cloud solution, with a multi-cloud GPU supply chain, fully managed platform, and one of the fastest LLM engines in the world. feel free to DM us via [email protected]

🚀 Excited to introduce the future of machine learning on Lepton AI! We've teamed up with Google's latest marvel, Gemma, to bring you an API that's as powerful as it is user-friendly. 🌟 Start exploring with Gemma today! #MachineLearning #AI #Gemma

Building an AI app has never been easier. Over the weekend, we built a demo for conversational search with <500 lines of python! Code to be open sourced soon as we clean up all "# temp scaffolds" stuff. (1/x)

Guess who got a spot on HuggingFace Space of the Week? It's us! Try tryemoji.com to checkout the emoji re-imagined with LCM 🚀🚀🚀 Bet you've never seen an emoji like this 😆 #huggingface #lcm #SDXLturbo #sdxl #imageGen

🚀 Lepton AI's model API now supports #NodeJS calls! 🖥️ Build smarter and faster with the power of AI in your favorite runtime. Dive in now! #LeptonAIUpdate

Lepton is created by a group of passionate open source contributors behind caffe, onnx, pytorch 1.0, etcd and more. We’ve open-sourced substantial portions of our stack, and will continue to open-source more in the upcoming months. Stay tuned!