Expand description
2026-09-25: DeepSeek-V4.1 full tiny-graph model: the config, weights, state
and trace types that forward (in model/forward.rs) fills. On the
fixture: embed, expand to hc_mult copies, the one-hot initial pre-mix,
engram before layers 1 and 4, six blocks with the delayed mixes,
collapse, final RMSNorm, f32 logits. The per-layer caches and the shared
slots carry across prefill and decode, so the golden’s captures for every
layer and the logits are checkable.
Owner: model-arch, DeepSeek-V4.1 reference. Invariants: none beyond the types.
Structs§
Functions§
- forward
- 2026-09-25: The forward for one chunk (
[seqlen]ids atstart_pos), text only.