Development build. This page describes main, not a released package. The latest published Lodestar.Embeddings is 0.7.0 — read its documentation.

Home › Embeddings › Tokenization

WordPieceVocabulary.GetHashCode

A hash consistent with that equality.

public int GetHashCode()

Returns — int, over the settings and the vocabulary's size rather than its every entry.

Example — equal vocabularies hash alike.

using System.Text;
using Lodestar.Embeddings.Persistence;
using Lodestar.Embeddings.Tokenization;

var bounds = new ArtifactLoadOptions();
byte[] file = Encoding.UTF8.GetBytes("[UNK]\ntoken\n##ize\ntext");

WordPieceVocabulary first = VocabTxtLoader.Load(
    new MemoryStream(file), bounds, unkToken: "[UNK]", continuationPrefix: "##", lowercase: true);
WordPieceVocabulary second = VocabTxtLoader.Load(
    new MemoryStream(file), bounds, unkToken: "[UNK]", continuationPrefix: "##", lowercase: true);

bool equal = first.Equals(second);  // => True
bool hashesAlike = first.GetHashCode() == second.GetHashCode();  // => True

Remarks — hashing every entry of a thirty-thousand-token vocabulary on every call is a cost nobody wants, so the hash is over the settings and the count. Two different vocabularies of the same size and settings collide, which is permitted: Equals is what decides, and it does read every entry.

Applies to — net10.0, netstandard2.0.

See also — WordPieceVocabulary.Equals.