Moritz

Munich today in the sunshine: Taking the @UnitreeRobotics dog for a walk.

Happy 25th birthday, TNG - coding since January 1st 2001!

Yesterday, the TNG's Chimera family of 671B child models, constructed from @deepseek_ai-parent LLMs (谢谢), processed in excess of 10 billion tokens/day on @openrouter for the first time. It is running on @chutes_ai backends.

Zen Driving Études after a long day of robotics work.

DeepSeek-TNG-R1T2-Chimera is currently the #1 trending text-generation model on @huggingface It got added to @openrouter a couple of hours ago, available for everyone to use. Thanks to @pingToven at @openrouter for adding the model and to @chutes_ai for providing the @nvidia H200s nodes.

Want to try out R1T2 Chimera? Here is the direct link to chutes: chutes.ai/app/chute/4fa0… Despite heavy attacks against its platform, @chutes_ai stabilised the deployment. Thx @jon_durbin

Today we release DeepSeek-TNG R1T2 Chimera. This new Chimera is a Tri-Mind Assembly-of-Experts model with three parents, namely R1-0528, R1 and V3-0324. R1T2 operates at a sweet spot in intelligence vs. output token length. It appears to be... * about 20% faster than R1, and more than twice as fast as R1-0528 * significantly more intelligent than R1 in benchmarks such as GPQA Diamond and AIME-24/25, albeit not quite on R1-0528 level * much more intelligent than our first R1T Chimera, and also think-token consistent, which is a major improvement We perceive it as generally well-behaved and a nice persona to talk to. The weights are on @huggingface under the MIT licence. We are looking forward to your experiments and feedback! Thanks to @deepseek_ai for giving their models to the world, to @chutes_ai and @openrouter for hosting R1T, to @WolframRvnwlf for benchmarking it, to @xlr8harder for beta-testing the new Chimera, and to @natolambert for constructive discussions at @aiDotEngineer.

Self-reflecting its own birthday...

Today we release DeepSeek-R1T-Chimera, an open weights model adding R1 reasoning to @deepseek_ai V3-0324 with a novel construction method. In benchmarks, it appears to be as smart as R1 but much faster, using 40% fewer output tokens. The Chimera is a child LLM, using V3s shared experts augmented with a custom merge of R1s and V3s routed experts. It is not a finetune or distillation, but constructed from neural network parts of both parent MoE models. A bit surprisingly, we did not detect defects of the hybrid child model. Instead, its reasoning and thinking processes appear to be more compact and orderly than the sometimes very long and wandering thoughts of the R1 parent model. Model weights are on @huggingface, just a little late for #ICLR2025. Kudos to @deepseek_ai for V3 and R1!

At TNG, we handle 5,000+ #LLM requests per hour and generate 10+ million tokens every day. Learn how our team optimizes inference serving for low-latency responses in high-traffic environments in the second article of our series on LLM performance: huggingface.co/blog/tngtech/l…

Load testing @nvidia RTX 5090 💯% GPU-utilization 87% vRAM-usage Burning 550 watts while running @Google gemma-3-27B at 105 tokens/s total throughput with @ollama Waiting for @vllm_project to ship a @Docker image compatible with the RTX 5090

Unlocking efficient unit test generation with Large Language Models. Discover how #LLMs can enhance developer productivity. Read the full article here: linkedin.com/pulse/leveragi…

Presenting Mixture of Tunable Experts (MoTE): Behavior Modification of DeepSeek-R1 at Inference Time - a method extending the MoE architecture of LLMs. By tuning 10 key experts, it enables meaningful and focused behavior changes on-the-fly. Mon 18h CET: linkedin.com/events/mixture…

Every company has their own optimized workflows, but not all can be expressed in #Jira out of the box. Remove that limiter with Workflow Enhancer, and access all fields in your transition element to craft the perfect #workflow for your work processes 👉 marketplace.atlassian.com/apps/575829/wo…

Are you subscribed to our TNG Apps YouTube channel yet? In our videos, you will find an overview of our popular apps for #Jira and #Confluence and gain new ideas on how to significantly enhance your #Atlassian experience. Check out the channel: @tngapps" target="_blank" rel="nofollow noopener">youtube.com/@tngapps

Join us at Atlas Camp '25 in Brussels on Feb. 3rd and 4th. Our colleagues Lina and Miroslava will share how they pioneered the migration of our Enterprise Survey app to the Cloud using @Atlassian's app migration platform for Forge. You can sign up here: events.atlassian.com/atlascamp25

Welcome to our dedicated space for news and announcements about our #tngapps for the #Atlassian ecosystem. Looking to improve #collaboration and #productivity on #Jira or #Confluence? Then we’ve got you covered. Check out appsbytng.com

Happy holidays from TNG! To create the cover for this year's Christmas cards, we used three different AI tools: #ChatGPT for a text-to-image prompt, #Midjourney to generate festive images and #RunwayML to animate the pictures. Our making-of: howto.christmascards.tngtech.com

It's pretty impressive to see this live. You can sense quite some tension in the air and heads burning 😅 #tngbtd @VincentKeymer04