Lodestar.Embeddings 0.4.0. This page is frozen at that release. Read the current documentation for what main says now. A link to a decision or a migration page follows main, and leaves the archive.

Reads the BPE model a tokenizer.json declares.

public static BpeVocabulary LoadBpe(Stream source, ArtifactLoadOptions options = null)
public static BpeVocabulary LoadBpe(string path, ArtifactLoadOptions options = null)

Parameters — source is a readable stream, never disposed here; path is the file to read. options bounds what will be accepted and defaults to ArtifactLoadOptions's own defaults.

Returns — BpeVocabulary, with the merge list, the split pattern and the byte-level flag all read from the file.

Exceptions — ArgumentNullException for a null source or path. InvalidDataException when the file declares a different model, declares a pipeline this package does not reproduce, or exceeds a bound in options — the message names what was refused and why.

Example — the preferred BPE route, and the refusal worth knowing about.

using Lodestar.Embeddings.Persistence;
using Lodestar.Embeddings.Tokenization;

BpeVocabulary vocab = TokenizerJsonLoader.LoadBpe("gpt2/tokenizer.json");

Remarks — Prefer this over BpeFilesLoader.Load whenever the checkpoint ships a tokenizer.json. It reads ignore_merges, the split pattern together with its behavior and invert flag, and the byte-level flag straight from the file — every one of which is a parameter the older two-file route asks a caller to get right.

Llama-2 and Mistral v0.1 are refused here by name. They are trained as SentencePiece BPE with a Metaspace pre-tokenizer and byte_fallback, a third pipeline distinct from both lineages this package implements. A real file of theirs declares model.type == "BPE", so this is the call that reaches byte_fallback, and the exception says what would go wrong: Python resolves an uncovered character into <0x..> byte pieces where this tokenizer emits the unknown piece, so loading it anyway would produce embeddings that do not match the model.

docs/decisions/0017-bpe-parity-scope.md has the parity scope — end to end for GPT-2 and the classic lineage, split-pattern only for Llama-3 and Qwen2 — and a known split divergence above the Basic Multilingual Plane.

Applies to — net10.0, netstandard2.0.

See also — BpeFilesLoader.Load, TokenizerJsonLoader.LoadBpeAsync, TokenizerJsonLoader.