Putting the ML in YAML. Eng @CoreWeave. Previously at @MicrosoftAI, @InflectionAI, @Cohere, @Color I like smash melee
Colorado, USA
- Tweets 3,401
- Following 145
- Followers 501
- Likes 12,199
Didn't call it a system one model. Ngmi
Introducing Julia-1:
Our first classification model that runs on almost anything.
Learn more 👇
supersoniclabs.ia.br/julia-1…
1
23
3,458
Imagine referring to rust as the slow version. How magical
While I was sleeping, Opus improved efficiency further. Now up to 22x faster at peak, 6x faster on mean, and up to 450x faster than the original Python implementation! Supports SSE2, AVX2, and AVX-512. Falls back to the slower Rust implementation otherwise.
194
Aaron retweeted
Opus 5.5 is:
- Smart
- Fast
- Fun to talk to
- Good at writing
- Friendly on rate limits
- Constantly surprising at just how much it is capable of
59
37
1,738
80,434
The amount of whiplash I've experienced switching back to team Claude with the release of Opus 5.5 should count as an injury in the work place. It's that good. And the usage limits are surreal.
1
142
My life for Peter
Congratulations to our Chief Technology Officer, Peter Salanki (@sorcer) named to the 2026 @TIME Executives of the Year: Tech and Data list.
Well deserved, Peter: bit.ly/4rqvTmx
2
6
305
Does anyone believe Boris when he says he uses this model for his daily driver? I would expect him to be wilding out on some internal model
Opus 5.5 is a really good model. It's been my daily driver the last few weeks.
We had Opus 5.5 and Fable 5.1 each port HAProxy from C to Rust. Both passed nearly all of HAProxy's tests, but Opus 5.5 finished in 9.5 hours compared to Fable 5.1's 12 hours, and for 51% less cost.
159
