🔥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

Introducing TxGemma, a family of open models specifically tailored for health settings, building on top of Gemma and Gemini. developers.googleblog.com/en/introducing…

Some nice quantization work makes our Gemma 3 model even more compact.

New Anthropic research: AI values in the wild. We want AI models to have well-aligned values. But how do we know what values they’re expressing in real-life conversations? We studied hundreds of thousands of anonymized conversations to find out.

Great to see Spanner recognized with this year's ACM SIGMOD Systems Award "for reimagining relational data management to enable serializability with external consistency at global scale". I was fortunate to be able to work on the first versions of Spanner with awesome colleagues

The Gemini API through our OpenAI compatibility layer now supports reasoning efforts: "low", "medium", "high", and "none"! You can hot swap to Gemini 2.5 Flash with 3 lines of code changed : )

Gemma just passed 150 million downloads and over 70k variants on Hugging Face🚀🚀🚀 What would you like to see in the next Gemma versions?

We're bringing the Unsloth magic to TTS and audio models! There are multiple free Colab notebooks with free GPUs for Whisper, Sesame, Orpheus, Spark, Llasa & Oute on our docs! docs.unsloth.ai/basics/text-to…

The Gemma team keeps shipping. In 6 months: - PaliGemma 2 - PaliGemma 2 Mix - Gemma 3 - ShieldGemma 2 - TxGemma - Gemma 3 QAT - Gemma 3n Preview - MedGemma Early - DolphinGemma - SignGemma And so much more to come! 🚀

Apache Spark 4.0 is out with some huge improvements across the board. SQL’s much more powerful, Spark Connect makes it easier to run apps, new languages and more. It’s amazing to see the community still growing fast and releasing over 5000 patches in 4.0. databricks.com/blog/introduci…

I'm doing a 3 hour advanced workshop at AI Engineer World's Fair 3rd June SF Marriott! Explore hacks to RL, GRPO, reward functions, kernel tricks, quantization & more! Also comment on any burning questions which I'll answer! Last year's workshop: youtu.be/pRM_P6UfdIc

See you all at 9am PT for my AI Engineer World's Fair workshop! I'll be at Foothill C. 👋 Excited to chat and meet! Also we'll be handing out limited edition Unsloth AI merch and stickers. If you can't make it today, we'll also be here throughout this week.

Nice to see folks studying biases in RLHF / preference tuning all the way down to the datasets. I think many of the biases are mostly irreducible human biases that can't be solved within current training regimes, just mitigated.

A strong agentic AI product built towards the vision of the CaptainAgent research (arxiv.org/abs/2405.19425), raising the power of a single agent to a new level using multi-agent architecture ⏬

Congratulations Mistral on launching their first reasoning model *and* writing an excellent report!

Q-learning is not yet scalable seohong.me/blog/q-learnin… I wrote a blog post about my thoughts on scalable RL algorithms. To be clear, I'm still highly optimistic about off-policy RL and Q-learning! I just think we haven't found the right solution yet (the post discusses why).

Feels good to be back coding! Just picked a fun one from my “someday” side project list and finally added a KV cache to the LLMs From Scratch repo: github.com/rasbt/LLMs-fro…

We made a complete Guide on Reinforcement Learning for LLMs! Learn about: • RL's goal & why it's key to building intelligent AI agents • Why o3, Claude 4 & R1 use RL • GRPO, RLHF, DPO, reward functions • Training your own local R1 model via Unsloth 🔗docs.unsloth.ai/basics/reinfor…

Gemma Developer Meetup at Google 🗓️ June 26th in San Francisco meetup.com/gemma-develope…

We're teaming up with Google for a Gemma developer meetup at Google's San Francisco office next Thursday, June 26! 🦥 • Join us & the Gemma team for live demos and talks • Unsloth new RL notebook & roadmap • Q&A + merch from us all RSVP required: lu.ma/gemma-unsloth

We're hosting an event on RL, GRPO, agents, LLM bugs & everything about Gemma 26th at Google's SF office! There are 3 Google DeepMind talks, special announcements & we're accepting 3 minute lightning talk proposals! Plus exclusive Unsloth merch! RSVP lu.ma/gemma-unsloth