in Runtime/Tokenizers/Tokenizers/Tokenizers.cs [53:68]
public TokenizerModel(JObject config, JObject tokenizerData)
{
Config = config;
TokenizerData = tokenizerData;
Vocab = new List<string>();
// Dict of tokens to ids Dict<string, number>
TokensToIds = new Dictionary<string, int>();
UnkTokenId = null;
UnkToken = null;
EndOfWordSuffix = null;
// Whether to fuse unknown tokens when encoding (default to false)
FuseUnk = false;
}