The AI development platform - From idea to AI, Lightning fast ??. Code together. Prototype. Train on GPUs. Scale. Serve. From your browser - with zero setup. AI Studio is your laptop on the cloud. Zero setup. Always ready. Persistent storage and environments. Code on CPU. Debug on GPU. Scale to multi-node. Run sweeps, jobs and more. Scale models with PyTorch Lightning, Fabric, Lit-GPT, torchmetrics and more.
Lightning AI的外部链接
US，NY，New York
It was great having you all Comet and Jayeeta Putatunda!
We had a great time connecting with the #GenAI community last week in New York! ?? Big thanks to Jayeeta Putatunda for her insightful talk, “Evaluating the Evaluators!”, where she dove into the complexities of assessing LLMs — covering everything from evaluation metrics to open-source initiatives. Also, huge shoutout to the Lightning AI team for hosting! ? If you’re in the NYC area, keep an eye out for our next in-person meetup ??
New from Lightning AI Studios ??more GPU machines now available for 50 - 80% cheaper! Choose between 1, 4 or 8 GPU instances. ??T4 x 1 from $0.41/hr ??L4 x 1 from $0.59/hr ??A10G x 1 from $0.74/hr ??L40S x 1 from $1.39/hr ??A100 x 8 from $13.48/hr ??H100 x 8 from $49.39/hr Lightning AI Studios give a local dev experience with power of 1,000s of cloud GPUs. Connect your local IDE, unlimited storage, multiplayer collaboration, to prototype, train, deploy, host AI web apps. Sign up, try them out, and?get 15 free credits every month! https://lightning.ai/
We're proud to announce that our CTO, Luca Antiga, is the first-ever Chair of the #PyTorch Foundation’s Technical Advisory Council! He and Vice Chair Jiong Gong bring extensive experience to guide the TAC in its mission to foster an open, diverse, and innovative PyTorch technical community. Learn more ? https://hubs.la/Q02SFVl90
Last week Luca Antiga and Thomas Viehmann explore clever optimizations of the Liger-Kernel team LinkedIn. Check out the full episode to see them use the Thunder compiler to add optimized Triton kernels to an existing Llama 3.2B implementation. See the full episode ?? https://lnkd.in/etQQhEkt And tune in to Livestream #11 tomorrow, Fri Oct 11 @ 11am ET, where they take a closer look at the resulting speedups ?? https://lnkd.in/ehM6ku5u
Today’s featured studio by Bhimraj Yadav is your guide to create your own chatbot in minutes with AI at Meta's LLaMA 3.2 Vision Multimodal LLM and deploy it with LitServe for lightning-fast performance. ? Tuned for both text and images ? Outperforms many open-source and closed alternatives ??LitServe optimized with OpenAI-compatible APIs, tool calling, and custom response formats Try it yourself ? https://lnkd.in/eauUCJBh
?? NYC's AI & ML devs! Join us & Comet tomorrow evening for a technical meetup! Hear from Jayeeta Putatunda on the challenges of evaluating LLMs and enjoy rooftop networking and refreshments. RSVP: https://lnkd.in/dtFQJuyk
Lightning AI转发了
Q4 is an ideal time to use remaining budget to run your AI/ML experiments. ?????? ??????? ??? Year-End Review: Gain insights to shape next year's strategy ??? Flexibility: Iterate on ideas w/o long-term commitments ??? Competitive Edge: Identify growth opportunities or areas to reduce spend ?? Have something in mind or just want to ideate on strategies??Let's talk
Congratulations to Geoffrey Hinton and John Hopfield on their 2024 Nobel Prize in Physics! Their work enables machine learning with artificial neural networks. This marks the first Nobel Prize in AI. What other breakthroughs do you think deserve recognition? Read more ? https://lnkd.in/dwQK3jnq
Deploy a conversational AI model using Jamba 1.5 with LitServe, ensuring scalability, privacy, and performance. ?? Long Context Handling: Industry-leading 256K context window, perfect for document summarization, complex analysis, and RAG workflows. ? Speed: Up to 2.5X faster on long contexts, the quickest across all size classes. ?? Quality: Jamba 1.5 Mini and Large models outperform even larger competitors on the?Arena Hard benchmark. ??? Multilingual: Supports multiple languages including?Spanish, German, Arabic, and more. Try it free ???https://lnkd.in/e5SSDrPZ
Say hello to the latest version of LitServe, v0.2.3—that gives you smoother deployment with automatic Dockerization and zero-latency logging API... and more: ???Dockerization ???Logger API ???Callback API ??Middleware More ?? https://lnkd.in/gbA4vvEk