The library documents itself · Neural tier

stzNeuralModel

21 own methods · 1 forms written with extensions, listed under their method · Also inherits from stzNeural (2)

the source: base/neural/stzNeuralModel.ring(opens in a new tab) · All classes · The area reference: Neural tier

Under each method, its extensions (Q CS XT Z ZZ IB W U): bold and ticked, it exists; struck, it does not. The extensions and what they mean

No method of this class is run on this page: the class reaches files, the network, processes, the clock, chance or sound, and an example run here would only be an example of its effects. Each method carries the explanation the library gives of itself.

Method

Explanation, as the source carries it

init

(no comment in the source: the name reads as a sentence) initno extension

LoadFrom

(Named LoadFrom, not Load -- "Load" collides with Haro's load keyword.)also written: Openno extension

LoadStatus

Why the last LoadFrom was refused (0 ok, -1 unreadable, -2 digest mismatch, -3 no recorded digest).no extension

IsLoaded

load / unload (runtime gguf)no extension

Unload

load / unload (runtime gguf)no extension

Path

load / unload (runtime gguf)no extension

Arch

(no comment in the source: the name reads as a sentence) archalso written: Architecturein the section architecture + hyperparametersno extension

EmbeddingDim

(no comment in the source: the name reads as a sentence) embedding dimin the section architecture + hyperparametersno extension

NumberOfLayers

(no comment in the source: the name reads as a sentence) number of layersin the section architecture + hyperparametersno extension

NumberOfHeads

(no comment in the source: the name reads as a sentence) number of headsin the section architecture + hyperparametersno extension

ContextLength

(no comment in the source: the name reads as a sentence) context lengthin the section architecture + hyperparametersno extension

VocabSize

(no comment in the source: the name reads as a sentence) vocab sizein the section architecture + hyperparametersno extension

NumberOfTensors

(no comment in the source: the name reads as a sentence) number of tensorsin the section architecture + hyperparametersno extension

Content

Content() = what the model IS: its architecture + hyperparameters as [key, value] data. Show() renders it (Softanza Show = visualize Content).also written: Infono extension

Show

(no comment in the source: the name reads as a sentence) showin the section architecture + hyperparametersno extension

EmbeddingOf

EmbeddingOf(cText) -- run the BERT forward pass (embeddings + N transformer layers + mean-pool + L2-normalize) and return the sentence-embedding vector as a list of EmbeddingDim() floats (DATA).also written: Embeddingno extension

Tokenize

WordPiece token ids for cText (with [CLS]..[SEP]) -- DATA.no extension

IsGenerative

TRUE if this model can GENERATE text (a causal decoder).no extension

Generate

Greedy generation from a raw prompt ("" when not generative).no extension

AnswerTo

Ask the instruct model a question (ChatML-wrapped, greedy).QCSXTZZZIBWU

SemanticSimilarityBetween

Cosine similarity of two texts' embeddings, in [-1, 1] (DATA). The vectors are already L2-normalized, so cosine = dot product.no extension