Lattice
Blog
Research notes from the Lattice project.
Quark on your Mac: the MLX build is here
A 1.7GB 4-bit MLX build of Quark, running fully on-device at ~100 tokens a second on Apple Silicon. How the conversion worked, and how to run it yourself.
Quark: a 1.5B model built from nothing
No Qwen under the hood. Custom tokenizer, custom architecture, ~2B pretraining tokens, then instruction tuning. The honest writeup of what from scratch training at 1.5B looks like.
Pulse vs Spark: identity in the weights, or in a system prompt?
Both are Qwen2.5-1.5B fine tunes. One knows it's Lattice only when told; the other has it baked in. Same knowledge, different honesty.
Lattice Mini: a 42M model built from scratch
Most small models are stamps on someone else's pretrained brain. Mini is the other kind, custom tokenizer, pretraining, and tuning, all from nothing.