🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Gemma 2 just got even better! 🚀 New Japanese-tuned 2B model AND a $150K Kaggle competition to build Gemma models for every language. Great to have Sundar Pichai here to share the excitement! Read more: goo.gle/Gemma4Japan #GemmaDeveloperDay

Congrats to my colleagues Demis Hassabis & John Jumper for winning the Nobel Prize in chemistry for their awesome AlphaFold work. AlphaFold has had a dramatic impact across many fields including medicine, drug development & more. And yes, this is my second Nobel-congratulating

Give a warm welcome to the @HuggingFace Ovis 1.6 Gemma 2 9B vision language model, one of the most recent additions to the #Gemmaverse. Check out this small but mighty vision model in this Gradio space for yourself → goo.gle/3YktqNB

This is a great opportunity! many people usually ask if there is an open position on my team and the answer now is: YES!

This Thursday I'm presenting at the AI Engineer London Meetup event I'll talk about Gemma models (of course) Come chat and have fun! lu.ma/scxafc58?tk=ym…

🔊✨ Speaker Announcement ✨🔊 We're excited to welcome Gus Martins(Gus (🤖🧠+🐍+🥑🗣️)), Google AI DevRel Engineer, to #DevFestLondon2024! Gus will be giving a workshop on #Gemma, introducing the world of open models and AI 🧠🪄 Learn how to harness Gemma's capabilities for text

Very good post!!! Gemma 2 2B showing great results (as expected) always make me happy! check out Andre's blog post and code, maybe add more models to the benchmark?

I'm very excited about this event! I'll present a Gemma workshop to give you a good start and help you with ideas! Come join us in person if you can (you won't regret!) but if you can't, you can follow the workshop online and participate in the hackathon from wherever you are!

mlx-vlm v0.1.4 is here 🎉 New models: - Google DeepMind Paligemma 2 Up next 🚧: - Refactoring Get started: > pip install -U mlx-vlm Please leave us a star and send a PR :)

we are so back! finally, PaliGemma2 for LaTeX OCR - used google/paligemma2-10b-pt-224 checkpoint - trained on A100 with 40GB VRAM - 4h of training on 20k training examples - QLoRA with 4bit quantization colab with complete fine-tuning code: colab.research.google.com/github/roboflo…

This is so cool! 💎 + 🎤 = 🤩 I liked how this is focused in the on-device story and I can see many interesting use cases when you can talk directly to the model

Gemma 2.0 27B is solid for code. It's definitely not as fast as Gemini 2.0 Flash, and not as good as 2.0 Pro, but definitely better than Llama 3.3 for my (heavily subjective) test suite and outputs.

This is cool! Gemma 2 still rocking and helping the GenAI field move forward! I wonder what will come next? 👀

Gemma 3 is here! Our new open models are incredibly efficient - the largest 27B model runs on just one H100 GPU. You'd need at least 10x the compute to get similar performance from other models ⬇️

🤯 Gemma 3's image analysis blew me away! Tested 2 ways to extract airplane registration numbers from photos with 12B model: 1️⃣ Gradio app w/API link (underrated feature IMO) + ZeroGPU infra on Hugging Face in Google Colab. Fast & free. 2️⃣ LM Studio server + local processing

See what’s new in Gemma 3, the latest generation of open models from Google with sizes ranging from 1B to 27B parameters. These versatile models run quickly and efficiently, and scale to developer needs to provide project flexibility.

Just announced new versions of Gemma 3 – the most capable model to run just one H100 GPU – can now run on just one *desktop* GPU! Our Quantization-Aware Training (QAT) method drastically brings down memory use while maintaining high quality. Excited to make Gemma 3 even more