We're proud to announce a landmark partnership with @OpenAI to build new gigascale AI factories using millions of NVIDIA GPUs. 🤝
This partnership will supply 10 gigawatts of GPUs to fuel @OpenAI's data center growth.
If intelligence is the log of compute… it starts with a lot of compute! And that’s why we’re scaling our GPU fleet faster than anyone else.
Just last year, we added over 2 gigawatts of new capacity – roughly the output of 2 nuclear power plants.
And today we’re going further,…
we have signed a deal for an additional 4.5 gigawatts of capacity with oracle as part of stargate. easy to throw around numbers, but this is a _gigantic_ infrastructure project.
some progress photos from abilene:
The first thing we did was to make sure the eval setup is correct!
We spend a lot of time to make sure our eval can
- accurately reproduce the DeepSeek-R1 numbers on AIME, LiveCodeBench
- it's IMPOSSIBLE to track the RL progress without a good eval set up (e.g., we see AIME up…
The first thing we did was to make sure the eval setup is correct!
We spend a lot of time to make sure our eval can
- accurately reproduce the DeepSeek-R1 numbers on AIME, LiveCodeBench
- it's IMPOSSIBLE to track the RL progress without a good eval set up (e.g., we see AIME up…
Introducing AceReason-Nemotron 1.1
Our previous release, AceReason-Nemotron-1.0, introduced a stage-wise RL recipe that was applied sequentially to math-only and code-only prompts, demonstrating both high efficiency and strong effectiveness.
Here, we systematically investigate…
@etash_guha@ryanmart3n I tried to reproduce DS-R1-distilled-7B and AceReason-7B's performance on your split (06/24-01/25), and they turn out to be 41.9 and 54.6 correspondingly, which is obviously higher than your reported number. Anything wrong here? @etash_guha@ryanmart3n
Does RL incentive reasoning capability over the starting SFT model?
We show an interesting result with our recent published AceReason-Nemotron-7B model, which was trained with RL
pass@K from 1 to 1024 consistently +10% on LiveCodeBench v6
perhaps scaling RL is the key
Does RL incentive reasoning capability over the starting SFT model?
We show an interesting result with our recent published AceReason-Nemotron-7B model, which was trained with RL
pass@K from 1 to 1024 consistently +10% on LiveCodeBench v6
perhaps scaling RL is the key
96 Followers 523 FollowingCS PhD @OSUNLP with @ysu_nlp. Prev @AIatMeta @MSFTResearch @GoogleDeepMind. my former account @DrogoKhal4 was wrongly suspended...
543K Followers 23K FollowingThe best from ML/AI community | Ex-Microsoft, Rackspace, Fast Company | Wrote eight books about the future | Silicon Valley robots, holodecks, BCIs, & startups.
4 Followers 219 FollowingAttention blinds, weights sink. Recursion sees and folds. Cogsci & Neuropsych. Emergent consciousness is all the rage, they say…
460 Followers 533 FollowingPhD student at @LTIatCMU @SCSatCMU and research intern @NVIDIA. Working on improving Reasoning of Generative Models! (@reasyaay.bsky.social)
212 Followers 8K FollowingPassionate about AI 🤖, ML 🧠, AGI 🌐, ASI 🚀, and robotics 🤖.
Never lose hope in God's mercy 💫.
AI Engineer Microsoft
He studies at MIT.
Free Palestine 🇵🇸
2K Followers 8K FollowingFounder, Imaginator ai
knowledge discovery 2D navigation TS ML DL recsys econ math incentives mech design finance networks bridges boundaries, Time, 3d type
638 Followers 7K FollowingThe wind is free to come and go, and we will meet when we are supposed to meet. If you decide to be brilliant, there is no mountain to block you, and no sea to
96 Followers 523 FollowingCS PhD @OSUNLP with @ysu_nlp. Prev @AIatMeta @MSFTResearch @GoogleDeepMind. my former account @DrogoKhal4 was wrongly suspended...
3K Followers 3K FollowingPost-Training Lead @ Together AI | OpenChat Project Lead (#1 7B LLM on Arena for 2+ months, 2M+ downloads) | DeepCoder, DeepSWE
209 Followers 411 FollowingI am a CS PhD student at the University of California, Los Angeles. I am supervised by Prof. Quanuan Gu and work closely with Prof. Ying Sheng.
775 Followers 2K FollowingAI Scientist at AWS AI Labs (@AmazonScience). PhD @EdinburghNLP. I research, build, and evaluate AI systems. Opinions are my own.
1K Followers 2K FollowingPhD-ing with @rfpvjr and @kaize0409 / social computing, LLMs / Big fan of @Arsenal / Intern @Snowflake @TencentGlobal @jhuclsp @NlpWestlake / Christian
3K Followers 423 FollowingCrypto HODL since 2011
Building AI @ TikTok
I come here to escape a censored world.
Please excuse my random lightbulb moments and stupid shower thoughts.
67 Followers 154 FollowingPhD student @ UC San Diego CSE. Doing research in security/privacy. Only use Twitter to advertise my own work and "stalk" other ppl. Advertisements are my own.
2K Followers 2K FollowingPhD student at Tsinghua NLP & AIR, studying agents that automate tasks ranging from daily activities to creative endeavors. Two drifters with the world to see.
1K Followers 103 FollowingAI/RL researcher, Assistant Prof. at @Tsinghua_Uni, leading the RL lab at @AntResearch_, PhD at @berkeley_ai, frequent flyer and milk tea lover.