@gabriberton I’ll need to try! Would need to rebuild with new maps, etc.
English
Jason Stock
568 posts
Jason Stock
@itsstock
PhD, Research Scientist @argonne
Colorado, USA Katılım Ağustos 2012
432 Takip Edilen721 Takipçiler
@itsstock This is a cool use case!
Doing the first 4 steps with one line of code with VisMatch should help :)
English
Generative Models via Drifting in MLX.
A minimal single file implementation showing how we evolve a pushforward distribution with a drifting field during training for single-step inference. No ODE/SDE solvers, adversarial loss, or complex nets.
code: github.com/stockeh/mlx-dr…
English
But… It’s too bad they kept changing big map and removed the API
github.com/stockeh/warzon…
English
Years ago I built a Warzone mapper to track player location:
extract minimap frames
detect SIFT features
KNN + Lowe ratio
estimate planar homography
warp to global coords
take centroids over time
connect into trajectory
Might be time to revisit the first 4 steps with this lib.
Gabriele Berton@gabriberton
VisMatch is on pypi! VisMatch is a wrapper for image matching models, like LightGlue, RoMa-v2, MASt3R, LoFTR, and 50+ more! It's literally as simple as: pip install vismatch vismatch-match --inputs img0 img1 --matcher choose_any To run image matching on any 2 images [1/4]
English
This site was really useful for toying with the LLM on my homelab -> canirun.ai
Haven't played with the different quantizations but will definitely give that a run this evening.
English
@BoulderCAST this feels particularly personal to northern colorado
English
Parts of the northern Front Range are blanketed with unbearably thick smoke Friday morning from large fires burning in Nebraska. This smoke will quickly get blown away by late morning #cowx
Nathan Booth@NaturisticNate
The smoke is so thick in Loveland this morning, It's hurting my lungs to breathe it in. Air quality is super unhealthy!! #cowx #colorado #fire #weather #Loveland #Denver #noco #foco @BianchiWeather @CReppWx @BoulderCAST @KylieBearseWX @kodythewxguy
English
@forgelab_ai @itsstock Please search issues, this is fixed and will be part of 1.3.1.
English
Ghostty 1.3 is now out! Scrollback search, native scrollbars, click-to-move cursor, rich clipboard copy, AppleScript, split drag/drop, Unicode 17 and international text improvements, massive performance improvements, and hundreds more changes. ghostty.org/docs/install/r…
English
In the machine learning space, we’re actively exploring generative diffusion geometry and interpretability!
Harness Labs@HarnessLabs
Anyone interested in: - Rust / Zig / C++ - Mathematics - Machine Learning - Scientific Computing Come join us here! discord.com/invite/qEwPr3G…
English
@awnihannun You’ve helped and inspired me and so many others in such a positive way! Excited to continue following your journey.
English
Today is my last day at Apple.
Building MLX with our amazing team and community has been an absolute pleasure.
It's still early days for AI on Apple silicon. Apple makes the best consumer hardware on the planet. There's so much potential for it to be the leading platform for AI. And I'm confident MLX will continue to have a big role in that.
To the future: MLX remains in the exceptionally capable hands of our team including @angeloskath, @zcbenz, @DiganiJagrit, @NasFilippova, @trebolloc (and others not on X). Follow them or @shshnkp for future updates.
English
@dvsaisurya I feel like @_arohan_ nailed your dissertation title then: “Once You Precondition, You Cannot Stop Preconditioning”
English
@dani_avila7 One thing is still missing from this.. native ghostty splits with claude teams rather than needing tmux.
English
@GregKamradt Myself and other close friends have also started to read more textbooks recently too!
Jason Stock@itsstock
Day 1 of reading Keenan Crane’s Discrete Differential Geometry book
English
As models got better at coding I found my appetite for technical textbooks went up, not down.
Since then I’m with Will, I care more about system design and good architecture choices than lower level
I imagine this too will change to managing tests and verifications
will brown@willccbb
with the latest models, i am now finding myself thinking about complex system design problems more, not less. the magnitude of what can be reasonably attempted is monumentally larger. you need to make sure you’re asking it to build the right thing.
English
@awnihannun Can we pause for a second and reflect on how insane this is to come from a single prompt on your Mac!?
English
Prompt and screen recording of the game:
mlx_lm.generate --model mlx_model --prompt "Write an HTML and JavaScript page implementing space invaders" -m 16384
English
GLM-5 runs with mlx-lm on a single 512GB M3 Ultra in Q4. It's quite good in my initial testing and pretty fast as well.
It generated a highly functional space invaders game using 7.1k tokens at 15.4 tok/s and 419GB memory.
Thanks to @ActuallyIsaak and @kernelpool for the port.
Z.ai@Zai_org
Introducing GLM-5: From Vibe Coding to Agentic Engineering GLM-5 is built for complex systems engineering and long-horizon agentic tasks. Compared to GLM-4.5, it scales from 355B params (32B active) to 744B (40B active), with pre-training data growing from 23T to 28.5T tokens. Try it now: chat.z.ai Weights: huggingface.co/zai-org/GLM-5 Tech Blog: z.ai/blog/glm-5 OpenRouter (Previously Pony Alpha): openrouter.ai/z-ai/glm-5 Rolling out from Coding Plan Max users: z.ai/subscribe
English
Day 1 of reading Keenan Crane’s Discrete Differential Geometry book
English