107 points jasondavies 33 minutes ago 9 comments
mattnewton 23 minutes ago | parent
From a research perspective, I'm even more interested in the varlooped variant they're releasing https://thinkingmachines.ai/model-card/tml-d24-varloop-resea... I'm not aware of any other working open source models like that.
kcb 14 minutes ago | parent
liuliu 9 minutes ago | parent
EDIT: by "open problems", I meant the problems I don't really formulate how to solve it first in my mind and then throw to it. I just describe what I want and let it to figure it out.
walrus01 17 minutes ago | parent
https://huggingface.co/thinkingmachines/Inkling-2-Mini
https://huggingface.co/unsloth/Inkling-2-Mini-GGUF
In terms of the active parameters and token/s on slow or less capable hardware it seems like it should be in the same general performance category as qwen 3.6 35B A3B.
neya 13 minutes ago | parent
christkv 12 minutes ago | parent
montroser 4 minutes ago | parent
Hope that folks find this model more amenable to local usage - the sizing was largely informed by what folks on this subreddit ask for the most haha!
There are two particularly interesting aspects of this model for local users:
We do our training end-to-end in nvfp4, and although we haven't exhaustively benchmarked this, folks like Unsloth have noticed that our models seem to respond very well to being quantized, maintaining 90% top-1 even at at 1-bit quantization.
This model is looped, which allows more depth at the same parameter count, which we particularly notice seems helpful for coding tasks. :)
This is our first model targeted at local users, so let us know how this model is and anything you want to see out of future iterations!
https://www.reddit.com/r/LocalLLaMA/comments/1y1diuk/thinkin...
jakswa 4 minutes ago | parent