hezar.models.speech_recognition.whisper.whisper_speech_recognition_config module

class hezar.models.speech_recognition.whisper.whisper_speech_recognition_config.WhisperSpeechRecognitionConfig(vocab_size: int = 51865, num_mel_bins: int = 80, encoder_layers: int = 6, encoder_attention_heads: int = 4, decoder_layers: int = 6, decoder_attention_heads: int = 4, num_hidden_layers: int = 12, decoder_ffn_dim: int = 1536, encoder_ffn_dim: int = 1536, encoder_layerdrop: float = 0.0, decoder_layerdrop: float = 0.0, decoder_start_token_id: int = 50257, use_cache: bool = False, sampling_rate: int = 16000, is_encoder_decoder: bool = True, activation_function: str = 'gelu', d_model: int = 256, dropout: float = 0.0, torch_dtype: str = 'float32', attention_dropout: float = 0.0, activation_dropout: float = 0.0, init_std: float = 0.02, scale_embedding: bool = False, max_source_positions: int = 1500, max_target_positions: int = 448, pad_token_id: int = 50256, bos_token_id: int = 50257, eos_token_id: int = 50256, suppress_tokens: List[int] = None, begin_suppress_tokens: List[int] = <factory>, use_weighted_layer_sum: bool = False, classifier_proj_size: int = 256, apply_spec_augment: bool = False, mask_time_prob: float = 0.05, mask_time_length: int = 10, mask_time_min_masks: int = 2, mask_feature_prob: float = 0.0, mask_feature_length: int = 10, mask_feature_min_masks: int = 0, max_new_tokens: int = 448, generation_config: dict | hezar.models.speech_recognition.whisper.whisper_speech_recognition_config.WhisperSpeechRecognitionGenerationConfig = None)

Bases: ModelConfig

activation_dropout: float = 0.0

activation_function: str = 'gelu'

apply_spec_augment: bool = False

attention_dropout: float = 0.0

begin_suppress_tokens: List[int]

bos_token_id: int = 50257

classifier_proj_size: int = 256

d_model: int = 256

decoder_attention_heads: int = 4

decoder_ffn_dim: int = 1536

decoder_layerdrop: float = 0.0

decoder_layers: int = 6

decoder_start_token_id: int = 50257

dropout: float = 0.0

encoder_attention_heads: int = 4

encoder_ffn_dim: int = 1536

encoder_layerdrop: float = 0.0

encoder_layers: int = 6

eos_token_id: int = 50256

generation_config: dict | WhisperSpeechRecognitionGenerationConfig = None

init_std: float = 0.02

is_encoder_decoder: bool = True

mask_feature_length: int = 10

mask_feature_min_masks: int = 0

mask_feature_prob: float = 0.0

mask_time_length: int = 10

mask_time_min_masks: int = 2

mask_time_prob: float = 0.05

max_new_tokens: int = 448

max_source_positions: int = 1500

max_target_positions: int = 448

name: str = 'whisper_speech_recognition'

num_hidden_layers: int = 12

num_mel_bins: int = 80

pad_token_id: int = 50256

sampling_rate: int = 16000

scale_embedding: bool = False

suppress_tokens: List[int] = None

torch_dtype: str = 'float32'

use_cache: bool = False

use_weighted_layer_sum: bool = False

vocab_size: int = 51865

class hezar.models.speech_recognition.whisper.whisper_speech_recognition_config.WhisperSpeechRecognitionGenerationConfig(alignment_heads: List[List[int]] = None, begin_suppress_tokens: List[int] = <factory>, bos_token_id: int = 50257, decoder_start_token_id: int = 50258, eos_token_id: int = 50257, forced_decoder_ids: List[List[int]] = <factory>, is_multilingual: bool = True, max_initial_timestamp_index: int = 50, max_length: int = 448, max_new_tokens: int = 448, no_timestamps_token_id: int = 50363, pad_token_id: int = 50257, prev_sot_token_id: int = 50361, return_timestamps: int = False, suppress_tokens: List[int] = None, task_to_id: dict[str, int] = <factory>)

Bases: object

alignment_heads: List[List[int]] = None

begin_suppress_tokens: List[int]

bos_token_id: int = 50257

decoder_start_token_id: int = 50258

dict()

eos_token_id: int = 50257

forced_decoder_ids: List[List[int]]

is_multilingual: bool = True

max_initial_timestamp_index: int = 50

max_length: int = 448

max_new_tokens: int = 448

no_timestamps_token_id: int = 50363

pad_token_id: int = 50257

prev_sot_token_id: int = 50361

return_timestamps: int = False

suppress_tokens: List[int] = None

task_to_id: dict[str, int]