Struct FallbackTokenizer 

pub struct FallbackTokenizer { /* private fields */ }

Fallback byte-level tokenizer.

Used when no proper tokenizer is available. Maps bytes directly to token IDs. This is NOT recommended for production use but ensures Hydra can always run.

Note: For Hydra inference, prefer HydraByteTokenizer which matches the training tokenizer exactly.

Implementations

impl FallbackTokenizer

pub fn new() -> Self

Create new fallback tokenizer.

pub fn with_vocab_size(vocab_size: usize) -> Self

Create fallback tokenizer that maps to a specific vocab size.

Token IDs will be byte % vocab_size to ensure they fit within bounds.

Trait Implementations

impl Clone for FallbackTokenizer

fn clone(&self) -> FallbackTokenizer

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FallbackTokenizer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for FallbackTokenizer

fn default() -> FallbackTokenizer

Returns the “default value” for a type.

impl HydraTokenizer for FallbackTokenizer

fn encode(&self, text: &str) -> Result<Vec<u32>>

Encode text to token IDs.

fn decode(&self, tokens: &[u32]) -> Result<String>

Decode token IDs back to text.

fn vocab_size(&self) -> usize

Get the vocabulary size.

fn tokenizer_type(&self) -> TokenizerType

Get the tokenizer type.

fn truncate(&self, tokens: Vec<u32>) -> Vec<u32>

Truncate tokens to maximum length for Hydra.

fn encode_for_hydra(&self, text: &str) -> Result<Vec<u32>>

Encode and truncate for Hydra input.