MCPcopy Create free account

hub / github.com/OpenMOSS/MOSS-TTS / functions

Functions1,298 in github.com/OpenMOSS/MOSS-TTS

⨍ Functions 1,298◇ Types & classes 178↳ Endpoints 16
↓ 3 callersMethod_drain_pending_tokens
(self)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:710
↓ 3 callersMethod_freeze_unused_qwen_embeddings
(self)
moss_tts_local/modeling_moss_tts.py:633
↓ 3 callersMethod_global_hidden_to_local
(self, hidden_states: torch.FloatTensor)
moss_tts_local_v1.5/modeling_moss_tts.py:193
↓ 3 callersFunction_gpu_gc
Force garbage collection and release GPU caches.
moss_tts_delay/llama_cpp/pipeline.py:79
↓ 3 callersMethod_local_text_candidate_ids
(self, device: torch.device)
moss_tts_local_v1.5/modeling_moss_tts.py:139
↓ 3 callersMethod_make_event
(self, chunk: np.ndarray, message_prefix: str)
moss_tts_realtime/app.py:284
↓ 3 callersFunction_normalize_path
(value: str, arg_name: str)
scripts/fuse_moss_tts_delay_with_codec.py:726
↓ 3 callersFunction_open_safe_reader
(path: Path)
scripts/fuse_moss_tts_delay_with_codec.py:68
↓ 3 callersFunction_require_namespace_path
(args: argparse.Namespace, attr_name: str)
scripts/fuse_moss_tts_delay_with_codec.py:741
↓ 3 callersFunction_require_namespace_str
( args: argparse.Namespace, attr_name: str, arg_name: str )
scripts/fuse_moss_tts_delay_with_codec.py:732
↓ 3 callersMethod_resolve_audio_items
( self, raw_audio_items: list[Any], n_vq: int, )
moss_tts_local_v1.5/processing_moss_tts.py:574
↓ 3 callersFunction_resolve_path
(path: str, base_dir: Optional[Path])
moss_tts_local_v1.5/finetuning/common.py:159
↓ 3 callersFunction_safe_float
(value: Any, *, default: float, minimum: float, maximum: float | None = None)
clis/moss_tts_local_v1.5_app.py:160
↓ 3 callersMethod_set_status
(self, *, state: str, error: str | None = None)
clis/moss_tts_local_v1.5_app.py:227
↓ 3 callersMethod_split_heads
(self, tensor: torch.Tensor)
moss_tts_local_v1.5/gpt2_decoder.py:131
↓ 3 callersFunction_stream_metrics
(*, now: Optional[float] = None)
moss_tts_local_v1.5/streaming.py:823
↓ 3 callersMethod_user_prompt_after_reference_ids
( self, language_code: object | None, prompt_fields: Optional[Dict[str, Any]], )
moss_tts_local_v1.5/processing_moss_tts.py:355
↓ 3 callersMethod_user_prompt_prefix_ids
(self)
moss_tts_local_v1.5/processing_moss_tts.py:348
↓ 3 callersMethodadd_noise
(self, original_samples, noise, timestep)
moss_soundeffect_v2/diffsynth/schedulers/flow_match.py:97
↓ 3 callersFunctionbasic_clean
(text)
moss_soundeffect_v2/diffsynth/prompters/wan_prompter.py:10
↓ 3 callersMethodbuild_assistant_message
( audio_codes_list: List[Union[str, os.PathLike, torch.Tensor]], content: str = AUDIO_PLACEHOL
moss_tts_local_v1.5/processing_moss_tts.py:304
↓ 3 callersMethodclear_voice_prompt
(self)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:584
↓ 3 callersMethodclose
(self)
moss_tts_delay/llama_cpp/backbone.py:230
↓ 3 callersMethoddecode
1. 这里不管怎样，都需要一个完整的 assistant generation ids; 2. 支持从任意位置进行截断；
moss_tts_local/processing_moss_tts.py:729
↓ 3 callersMethoddecode
1. 这里不管怎样，都需要一个完整的 assistant generation ids; 2. 支持从任意位置进行截断；
moss_tts_delay/processing_moss_tts.py:731
↓ 3 callersMethoddecode
(self, output: Any, *, return_stereo: bool = True)
moss_tts_local_v1.5/processing_moss_tts.py:755
↓ 3 callersFunctiondecode_audio_frames
( audio_frames: list[torch.Tensor], decoder: AudioStreamDecoder, codebook_size: int, audio_eos
moss_tts_realtime/example_multiturn_stream_to_tts.py:66
↓ 3 callersFunctiondecode_audio_frames
( audio_frames: list[torch.Tensor], decoder: AudioStreamDecoder, codebook_size: int, audio_eos
moss_tts_realtime/example_llm_stream_to_tts.py:66
↓ 3 callersMethoddecode_codes
(self, codes: torch.LongTensor)
moss_tts_local_v1.5/streaming.py:596
↓ 3 callersMethodencode_audios_from_path
( self, wav_path_list: Union[str, List[str]], n_vq: Optional[int] = None )
moss_tts_local/processing_moss_tts.py:872
↓ 3 callersMethodencode_audios_from_path
( self, wav_path_list: Union[str, List[str]], n_vq: Optional[int] = None )
moss_tts_delay/processing_moss_tts.py:874
↓ 3 callersMethodencode_audios_from_wav
( self, wav_list: List[torch.Tensor], sampling_rate: int, n_vq: Optional[int]
moss_tts_local/processing_moss_tts.py:796
↓ 3 callersFunctiongroup_weight_map_by_shard
(weight_map: dict[str, str])
scripts/fuse_moss_tts_delay_with_codec.py:213
↓ 3 callersMethodinitialize_local_text_lm_head_from_text_lm_head
(self)
moss_tts_local_v1.5/modeling_moss_tts.py:149
↓ 3 callersFunctionload_jsonl
(path: str | Path)
moss_tts_realtime/finetuning/common.py:21
↓ 3 callersFunctionload_jsonl
(path: str | Path)
moss_tts_local/finetuning/common.py:21
↓ 3 callersFunctionload_jsonl
(path: str | Path)
moss_tts_delay/finetuning/common.py:21
↓ 3 callersFunctionload_jsonl
(path: str | Path)
moss_tts_local_v1.5/finetuning/common.py:14
↓ 3 callersFunctionloudness_normalize
( wav: np.ndarray, target_dbfs: float = -20.0, gain_range: tuple[float, float] = (-3.0, 3.0), )
moss_tts_delay/llama_cpp/pipeline.py:55
↓ 3 callersFunctionnormalize_tts_text
对 TTS 输入做鲁棒性正则化。
moss_tts_delay/tts_robust_normalizer_single_script.py:85
↓ 3 callersFunctionprecompute_freqs_cis
(dim: int, end: int = 1024, theta: float = 10000.0)
moss_soundeffect_v2/diffsynth/models/wan_video_dit.py:95
↓ 3 callersMethodpush_text
(self, text_fragment: str)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:622
↓ 3 callersFunctionsample_token
( logits, prev_tokens: Optional[torch.LongTensor] = None, repetition_penalty: float = 1.0, top
moss_tts_delay/inference_utils.py:111
↓ 3 callersFunctionsample_token
Sample tokens from logits with optional filtering and repetition penalty.
moss_tts_delay/llama_cpp/sampling.py:104
↓ 3 callersMethodsave_model
(self, accelerator, model, file_name)
moss_soundeffect_v2/diffsynth/trainers/utils.py:806
↓ 3 callersMethodset_timesteps
(self, num_inference_steps=100, denoising_strength=1.0, training=False, shift=None, dynamic_shift_len=None)
moss_soundeffect_v2/diffsynth/schedulers/flow_match.py:34
↓ 3 callersFunctionsinusoidal_embedding_1d
(dim, position)
moss_soundeffect_v2/diffsynth/models/wan_audio_dit.py:18
↓ 3 callersMethodsnapshot
(self)
clis/moss_tts_local_v1.5_app.py:343
↓ 3 callersMethodsnapshot
(self)
moss_tts_realtime/app.py:675
↓ 3 callersMethodtie_weights
(self, *args, **kwargs)
moss_tts_local_v1.5/modeling_moss_tts.py:114
↓ 3 callersMethodto_dict
(self)
moss_tts_delay/processing_moss_tts.py:50
↓ 3 callersMethodtransformer
(self)
moss_soundeffect_v2/pipeline_moss_soundeffect.py:56
↓ 3 callersFunctionwrite_sorted_json
(path: Path, payload: Mapping[str, object])
scripts/fuse_moss_tts_delay_with_codec.py:96
↓ 2 callersMethod__exit__
(self, *args)
moss_tts_delay/llama_cpp/pipeline.py:825
↓ 2 callersMethod__init__
(self, in_dim, out_dim, has_pos_emb=False)
moss_soundeffect_v2/diffsynth/models/wan_audio_dit.py:58
↓ 2 callersMethod__init__
(self, entry)
moss_soundeffect_v2/diffsynth/models/wan_video_camera_controller.py:80
↓ 2 callersMethod_apply_crossfade
(self, wav: torch.Tensor, final_chunk: bool = False)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:828
↓ 2 callersMethod_assert_fixed_nq
(self, n_vq: Optional[int])
moss_tts_local_v1.5/processing_moss_tts.py:310
↓ 2 callersMethod_build_audio_rows
(self, audio_tokens: torch.Tensor, slot_token_id: int)
moss_tts_local_v1.5/processing_moss_tts.py:336
↓ 2 callersFunction_build_frame_decoder
( codec, inferencer: MossTTSRealtimeInference, device: torch.device, *, chunk_frames: int,
moss_tts_realtime/app.py:358
↓ 2 callersMethod_build_generation_row
( self, batch_size: int, device: torch.device, audio_token_ids: torch.LongTens
moss_tts_local_v1.5/modeling_moss_tts.py:446
↓ 2 callersMethod_build_lm_heads
(config: PipelineConfig)
moss_tts_delay/llama_cpp/pipeline.py:421
↓ 2 callersFunction_build_streaming_session
( model: MossTTSRealtime, tokenizer, processor: MossTTSRealtimeProcessor, codec, *, ma
moss_tts_realtime/app.py:324
↓ 2 callersMethod_build_text_only_turn_input
( processor: MossTTSRealtimeProcessor, user_text: str, prompt_tokens: np.ndarray | Non
moss_tts_realtime/app.py:519
↓ 2 callersMethod_causal_attention_mask
( self, attention_mask: Optional[torch.Tensor], query_length: int, key_length:
moss_tts_local_v1.5/gpt2_decoder.py:149
↓ 2 callersMethod_causal_attention_mask
( self, attention_mask: Optional[torch.Tensor], query_length: int, key_length:
moss_tts_local_v1.5/qwen3_decoder.py:148
↓ 2 callersMethod_compute_inv_freq
(self, device: Optional[torch.device] = None)
moss_tts_local_v1.5/gpt2_decoder.py:49
↓ 2 callersMethod_compute_inv_freq
(self, device: Optional[torch.device] = None)
moss_tts_local_v1.5/qwen3_decoder.py:60
↓ 2 callersMethod_consume_frames
(self, num_frames: int)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:787
↓ 2 callersMethod_convert_token_to_id
(self, token: str)
moss_tts_realtime/mossttsrealtime/processing_mossttsrealtime.py:68
↓ 2 callersMethod_decode
(self, tokens: torch.Tensor)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:803
↓ 2 callersFunction_decode_budget_from_stream_state
( *, lead_seconds: float, default_budget: int = 0, first_decode_submitted: bool = False, f
moss_tts_local_v1.5/streaming.py:659
↓ 2 callersFunction_decode_reference_path
(path: str)
clis/moss_tts_local_v1.5_app.py:171
↓ 2 callersMethod_encode_audio_tokens
( self, path: Path, codec, device: torch.device, chunk_duration: float
moss_tts_realtime/app.py:484
↓ 2 callersMethod_encode_audio_tokens
( self, path: Path, codec, device: torch.device, chunk_duration: float
moss_tts_realtime/fast_api.py:357
↓ 2 callersFunction_ensure_config
(cfg: Any, cls: type[PretrainedConfig])
moss_tts_realtime/mossttsrealtime/configuration_mossttsrealtime.py:25
↓ 2 callersFunction_extract_codes
(encode_result)
moss_tts_realtime/example_multiturn_stream_to_tts.py:25
↓ 2 callersFunction_generation_lead_seconds
(*, now: Optional[float] = None)
moss_tts_local_v1.5/streaming.py:810
↓ 2 callersMethod_get_audio_tokenizer_device
Best-effort device inference for `self.audio_tokenizer`. Notes: - Old TAC wrapper exposed `.device`, but standard `torch.nn.Module` d
moss_tts_local/processing_moss_tts.py:768
↓ 2 callersMethod_get_audio_tokenizer_device
Best-effort device inference for `self.audio_tokenizer`. Notes: - Old TAC wrapper exposed `.device`, but standard `torch.nn.Module` d
moss_tts_delay/processing_moss_tts.py:770
↓ 2 callersMethod_get_audio_tokenizer_device
(self)
moss_tts_local_v1.5/processing_moss_tts.py:788
↓ 2 callersFunction_init_nvml
()
moss_tts_delay/llama_cpp/gpu_monitor.py:41
↓ 2 callersFunction_is_static_kv_cache_layer
(layer_past: object)
moss_tts_local_v1.5/gpt2_decoder.py:36
↓ 2 callersFunction_iter_shard_tensors
( model_dir: Path, shard_groups: dict[str, list[str]], output_name_builder: Callable[[str], str],
scripts/fuse_moss_tts_delay_with_codec.py:264
↓ 2 callersFunction_load_audio
(path: Path, target_sample_rate: int = 24000)
moss_tts_realtime/example_multiturn_stream_to_tts.py:29
↓ 2 callersMethod_load_idx_if_needed
(self, shard_id: int)
moss_soundeffect_v2/diffsynth/trainers/cache_shards.py:229
↓ 2 callersMethod_lookup
(self, ids: np.ndarray)
moss_tts_delay/llama_cpp/embedding.py:70
↓ 2 callersMethod_mark_started
(self)
moss_tts_realtime/fast_api.py:126
↓ 2 callersFunction_normalize_attention_implementation
(value: Optional[str], default: str = "flash_attention_2")
moss_tts_local_v1.5/configuration_moss_tts.py:16
↓ 2 callersMethod_normalize_audio_tokens
(self, audio_tokens: np.ndarray | Iterable)
moss_tts_realtime/mossttsrealtime/processing_mossttsrealtime.py:90
↓ 2 callersFunction_normalize_spaces
(text: str)
moss_tts_delay/tts_robust_normalizer_single_script.py:163
↓ 2 callersMethod_overlap_samples
(self, wav: torch.Tensor)
moss_tts_realtime/mossttsrealtime/streaming_mossttsrealtime.py:853
↓ 2 callersMethod_packed_segments_from_cu_seqlens
( cu_seqlens: torch.Tensor, num_sequences: Optional[torch.Tensor], device: torch.devic
moss_tts_local_v1.5/gpt2_decoder.py:469
↓ 2 callersMethod_packed_token_indices
( batch_ids: torch.Tensor, starts: torch.Tensor, lengths: torch.Tensor, seq_le
moss_tts_local_v1.5/gpt2_decoder.py:519
↓ 2 callersMethod_pad
(self, input_ids_list: List[torch.Tensor])
moss_tts_local/processing_moss_tts.py:434
↓ 2 callersMethod_pad
(self, input_ids_list: List[torch.Tensor])
moss_tts_delay/processing_moss_tts.py:430
↓ 2 callersMethod_pad
(self, input_ids_list: list[torch.Tensor])
moss_tts_local_v1.5/processing_moss_tts.py:600
↓ 2 callersFunction_pad_or_trim_channels
(codes, pad_value)
moss_tts_realtime/finetuning/dataset.py:52
↓ 2 callersMethod_prompt_fields_from_user_message
(self, message: Dict[str, Any])
moss_tts_local_v1.5/processing_moss_tts.py:376
↓ 2 callersFunction_require_namespace_bool
(args: argparse.Namespace, attr_name: str)
scripts/fuse_moss_tts_delay_with_codec.py:748
← previousnext →101–200 of 1,298, ranked by callers