ReadonlyappendKnown tokens: this graph declares no committed spans, so they read as a window.
ReadonlycacheOne cache per layer, sharing one mask memo (create_attention_mask's one
mask per forward, built by the caches). Their TurboQuant read is dense
(denseTurboQuantKernels): the read decodes keys and values, then runs the
fused SDPA.
ReadonlyconfigReadonlydecodeOne new token per sequence: ids [sequences, 1].
ReadonlyembedUnscaled token embeddings, for media prompts that splice embeddings.
ReadonlyfinalReadonlylayersReadonlylmUntied lm_head (tie_word_embeddings=false for this family).
ReadonlylogitsThe one output head at every width.
ReadonlyloraReadonlymultiReadonlymultiReadonlyprefixReadonlypromptOne whole chunk of a prompt; rows are sequences times the chunk size. The GPU is this graph's one unit.
ReadonlypromptOptional chunk bindings when a file distinguishes chunks from prompt tails.
ReadonlysequenceFull-sequence scoring operations, indexed by total token rows.
ReadonlyverifyA speculative window of depth + 1 positions per sequence. This graph
declares no hidden-layer taps, so capture is always null.
ReadonlyweightsAn ordinary autoregressive graph: every method over every cache lego.
Builds no constants and opens no weights: nothing to release.
Protectedforward
The graph's operations and construction facts. Pooled embedding support is declared by its operation alone. Other operations accompany the promises in
graphCapabilities; composition checks their pairing when it binds (declaredGraph).