Byte Level Bpe Tokenizer#

class ByteLevelBpeTokenizer#

Public Functions

SafetyStatus declareResources(
std::filesystem::path const &artifact,
ProcessingLimits const &limits,
ByteLevelBpeTokenSpec const *requiredTokens,
std::size_t requiredTokenCount,
ResourceDeclaration &declaration
) noexcept#
SafetyStatus initialize(PreparedResources const &prepared) noexcept#
SafetyStatus encode(
TextView text,
int32_t *output,
std::size_t capacity,
std::size_t &outputSize
) noexcept#
SafetyStatus decode(
TokenIdView tokens,
TokenIdView stopTokens,
MutableTextBuffer output,
std::size_t &outputSize
) const noexcept#
int32_t specialTokenId(std::string_view token) const noexcept#
bool hasTokenId(int32_t id) const noexcept#
SafetyStatus deinitialize() noexcept#
struct ByteLevelBpeTokenSpec#

Public Members

std::string_view bytes = {}#
int32_t id = {-1}#
bool special = {false}#