6 projects
bantamtrain
Bantam: train small language models and weather models faster and cheaper on limited compute - Zoom Training, Muon, 8-bit/bf16 low-memory training, int8/int4 weights, compressed token and weather-grid stores
helix-lm
HELIX, ANANSI and TriStream-2: linear-scaling sequence architectures with exact long-range recall, the engine that trains them (Muon, muP, growth), and a music stack (analysis, frisson critic, conductor, separation, singing voice conversion)
naturev1
Probabilistic weather and hurricane model: geometric-attention backbone, tropical-cyclone physics that beats extrapolation, a satellite-fed particle swarm with Caribbean parish impacts, multi-teacher distillation, data assimilation
ihelix
One architecture for fields on any geometry: spheres, grids, volumes, periodic boxes and point clouds
recreator
Rebuild a pretrained transformer as a HELIX model without pretraining it again.
nanofold
FOLD compression: fit large models into small VRAM at near-lossless quality
