The official source reports this update: b10776: model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support. model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support ( #25444 ) hparams: add per-layer n_ff_exp/n_expert_used arrays with scalar-or-array loading G1/G2 infrastructure for variable-per-layer expert FFN size and top-k routing (required for Puzzle-75B which has 5 distinct n_ff_exp values and 7 top-k values across its 40 MoE layers). Design: rename scalar members to _impl suffix (following existing convention), add LLAMA_MAX_LAYERS arrays, add n_ff_exp(il)/n_expert_used(il) accessors with scalar fallback.
Canonical AI event
NVIDIA Nemotron-3-Puzzle-75B-A9B now supported with per-layer expert routing.
NVIDIA's Nemotron-3-Puzzle-75B-A9B model now supports per-layer expert routing. The update includes handling 5 distinct feed-forward sizes and 7 top-k values across its 40 layers.
4 Sept 20261 verified claims1 sources1 observations
This official update documents a development concerning b10776: model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support. Its practical significance depends on the scope and evidence stated by the source.
Read the official source update and verify its stated scope, evidence and timing before acting on it.
Connected knowledge
Entities affected by this event
Evidence trail
Sources behind the event
Editorial presentation
Open story