espnet2.sds.end_to_end.mini_omni.litgpt.model.GPT

source

class espnet2.sds.end_to_end.mini_omni.litgpt.model.GPT(config: Config)

Bases: Module

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

clear_kv_cache() → None

concat_whisper_feat(audio_feature, input_ids, T, task)

forward(audio_features: Tensor, input_ids: Tensor, input_pos: Tensor | None = None, whisper_lens: list | None = None, task: str | None = None) → Tensor

Define 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.

classmethod from_name(name: str, **kwargs: Any) → Self

property max_seq_length : int

reset_parameters() → None

rope_cache(device: device | None = None) → Tuple[Tensor, Tensor]

set_kv_cache(batch_size: int, rope_cache_length: int | None = None, device: device | None = None, dtype: dtype | None = None) → None