Repository navigation

Nemotron 3.5 Bolt 74B-A9B on CUDA: puzzle layers, LatentMoE, ggml IQ2_XXS/Q8_0 weights, image input, host prompt states (on #853) - #897

Closed

BobClawblaw wants to merge 12 commits into

ashhart:mainfrom

BobClawblaw:nemotron-bolt

Closed

Nemotron 3.5 Bolt 74B-A9B on CUDA: puzzle layers, LatentMoE, ggml IQ2_XXS/Q8_0 weights, image input, host prompt states (on #853)#897
BobClawblaw wants to merge 12 commits into
ashhart:mainfrom
BobClawblaw:nemotron-bolt

Commits on Oct 14, 2026
- lane core and native CUDA host for two-rank families with a vision tower: batched prompt passes (prefill_many, addStreams), a backend's refusals for a memory guard, follow() for a tensor-parallel r…
  BobClawblaw
  committed
  39de039
Commits on Oct 15, 2026