espnet2.speechlm.tokenizer.abs_tokenizer.AbsTokenizer

source

class espnet2.speechlm.tokenizer.abs_tokenizer.AbsTokenizer(*args, **kwargs)

Bases: Module, ABC

The abstract tokenizer class for SpeechLM. The main objective of this module is to transform the LM-generated tokens into the corresponding targets. E.g., Speech Codec codes -> waveform BPE tokens -> text …

Initializes internal Module state, shared by both nn.Module and ScriptModule.

abstract forward(tokens: Tensor) → Any

Defines the computation performed at every call.

Should be overridden by all subclasses.

NOTE

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.