Référence · Langage et intelligence
Palier neuronal
20 fonctions dans 7 classes, groupées par ce qu'elles font, chacune avec l'explication que la bibliothèque donne d'elle-même.
Le récit de ce domaine, pour un lecteur général : ce qu'un artisan en fait, un exemple exécuté et ses couloirs notés, est sur sa page de la Plateforme.
Les fonctions, par ce qu'elles font
Tester et vérifier
- IsConstrainingDecoding Will the NEXT live call be grammar-constrained? Ask this rather than assuming: a nested structure, an off switch, or a build without the sampler rung all answer 0, and WhyNotConstrained() says which. dans stzLLMFunction
- Check The family's rule face: run me, hand me back the findings. dans stzOutputSchema
- IsDecodingConstrained Does anything actually CONSTRAIN DECODING with this grammar? Ask before reporting an answer as grammar-constrained. It answered 0 for as long as that was true; gbnf_machine.zig is the rung underneath it, so it answers 1 since 2026-08-20 -- the ENGINE is the authority on that number, never this comment. dans stzOutputSchema
- IsExpressibleAsGrammar TRUE when this declaration can be expressed as a grammar at all. dans stzOutputSchema
- IsLoaded load / unload (runtime gguf) dans stzNeuralModel
- IsGenerative TRUE if this model can GENERATE text (a causal decoder). dans stzNeuralModel
- IsReady stateless facade over the ggml inference backend. TRUE if the ggml backend is compiled in and executes. dans stzNeuralEngine
Trouver
- LastFindings The findings the LAST validation produced, in the family's unified shape -- so a refusal can be handed to stzRuleReport.Ingest() and stand in the same CI gate as every other rule in the library. dans stzLLMFunction
- SearchByVectorXT The same search from a query embedding the caller already holds (the companion of AddEmbedded). ONE search path serves both doors: the route -- CPU resident dataset or GPU resident corpus -- is decided here and nowhere else. dans stzSemanticIndex
Compter et mesurer
- Count feeding the index dans stzSemanticIndex
Remplacer et modifier
- SetRetrySampling The sampling a RETRY uses. Attempt 1 never sees these -- it is greedy. dans stzLLMFunction
Ajouter
- AddGolden golden sets dans stzLLMFunction
- AddText Embeds at ADD time (the model must be loaded), so a bad text fails loudly here and not in the middle of a search. dans stzSemanticIndex
- AddTexts feeding the index dans stzSemanticIndex
- AddEmbedded Index a text with an embedding the CALLER already has -- another model, a cache, a wire -- so the index never insists on embedding it again. The vector must be L2-normalised (the score formula assumes it) and of the index's dimension. dans stzSemanticIndex
Convertir
- ToGBNF The GBNF text for this declaration, or a raise naming the construct that cannot be expressed. Nested structures are REFUSED rather than flattened: a grammar that accepted what this schema rejects would put the two layers into disagreement. dans stzOutputSchema
- ParseOutput RAW MODEL TEXT to a validated whole, or to a refusal that names the rule it failed. [ :ok, :value, :findings, :shape ] dans stzOutputSchema
Lire et écrire
- LoadFrom (Named LoadFrom, not Load -- "Load" collides with Haro's load keyword.) dans stzNeuralModel
- LoadStatus Why the last LoadFrom was refused (0 ok, -1 unreadable, -2 digest mismatch, -3 no recorded digest). dans stzNeuralModel
Parcourir et appliquer
- RunGoldens Goldens hold STRUCTURES as readily as scalars now. Two things had to change for that, and both were real defects rather than gaps: Haro's own `=` answers 0 for two identical lists, so a structured golden could never have passed; and a failing structured case that reports only "expected / got" is unreadable, so the failure now carries the FIELD that moved. dans stzLLMFunction
Les explications sont celles que la bibliothèque donne d'elle-même, en anglais, lues dans ses commentaires de documentation le 2026-10-01 04:14, from the library at commit 0e72e2e2c, by the library's own stzSelfDoc over 635 class sources.
Les classes de ce domaine
- stzLLMFunction · 31 méthodes
- stzOutputSchema · 28 méthodes
- stzNeuralModel · 21 méthodes
- stzSemanticIndex · 13 méthodes
- stzNeuralChat · 12 méthodes
- stzNeuralEngine · 6 méthodes
- stzNeural · 2 méthodes
Les narrations sur ce domaine
- A Guided Path to Problem-Solving: The Softanza Mental Model(s'ouvre dans un nouvel onglet)
- Reaxis: A Natural Model for Reactive Programming in Haro, by Softanza(s'ouvre dans un nouvel onglet)
- stzGraph — Modeling Knowledge Through Structure(s'ouvre dans un nouvel onglet)
- **Time Modeling in Softanza using `stzTimeLine`**(s'ouvre dans un nouvel onglet)

