Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/debpalash/VoiceStudio
/ functions
Functions
15,058 in github.com/debpalash/VoiceStudio
⨍
Functions
15,058
◇
Types & classes
1,240
↳
Endpoints
303
↓ 3 callers
Function
_bridge_request
The exact request shape the upstream integration sends.
tests/test_pyvideotrans_contract.py:75
↓ 3 callers
Function
_build
(gender, age, pitch, *, accent=None, dialect=None, whisper=False, use_case, name, icon, language, s
backend/core/archetypes.py:202
↓ 3 callers
Function
_burn_subs_allowed
Subtitle burn-in combined with video retime. Allowed for ``smart_fit`` (fitted cue records exist, and the burn pass runs AFTER the retime gra
backend/api/routers/dub_export.py:522
↓ 3 callers
Method
_cancel_active_work
Cancel retained tasks without permanently disabling reconnect.
backend/worker/transport/client.py:402
↓ 3 callers
Function
_candidate_roots
Source roots to probe, most-specific first. ``OMNIVOICE_PROJECT_ROOT`` lets the launcher point at the staged project dir explicitly; otherwis
backend/core/omnivoice_path.py:27
↓ 3 callers
Function
_capabilities
(resident: bool = False)
tests/test_worker_transport.py:67
↓ 3 callers
Function
_capability_payload_allowed
Bound model metadata before converting it into Python containers.
backend/worker/transport/server.py:321
↓ 3 callers
Function
_caps
(family, *, notes=())
tests/test_routing_redaction.py:17
↓ 3 callers
Function
_carry_srt_voice_metadata
Replace subtitle content while retaining the source cast assignment.
backend/api/routers/dub_core.py:170
↓ 3 callers
Function
_cinematic_budget
Overall wall-clock cap for a whole cinematic/autofit refine pass (seconds). Unfinished segments degrade to their literal (Fast) translation once h
backend/services/translator.py:241
↓ 3 callers
Function
_cli
()
backend/services/tailscale.py:22
↓ 3 callers
Function
_client
The shared httpx client, unless a caller (or a test) supplied one.
backend/services/gallery.py:355
↓ 3 callers
Function
_client
(module_name: str, attr: str = "router")
backend/tests/test_asr_request_path_degrades.py:124
↓ 3 callers
Function
_client
(app)
tests/backend/test_perf_settings.py:45
↓ 3 callers
Function
_collect
(cmd)
backend/tests/test_stream_stderr_sync_fallback.py:94
↓ 3 callers
Function
_cols
(db_path, table="voice_profiles")
tests/test_db_schema_reconcile.py:15
↓ 3 callers
Function
_common_props
()
backend/core/analytics.py:320
↓ 3 callers
Function
_concrete
A minimal concrete TTSBackend subclass (satisfies the ABC) that inherits the real default unload() under test.
tests/test_engine_memory.py:44
↓ 3 callers
Function
_connect
(path)
tests/test_generation_history_self_heal.py:20
↓ 3 callers
Function
_consume_native_save
(authorization: str)
backend/api/routers/dub_export.py:174
↓ 3 callers
Function
_contained_path
(root, value, *, detail="Invalid file path")
backend/api/routers/marketplace.py:61
↓ 3 callers
Function
_ctranslate2_cudnn_ok
Availability gate for the two CTranslate2 engines (WhisperX, faster-whisper). Importing them proves nothing about cuDNN 8: CTranslate2 only reach
backend/services/asr_backend.py:209
↓ 3 callers
Function
_dc
Resolve the app module at call time. Module-level imports of app modules go stale under sys.modules pollution from other suites (the `tests/*
tests/test_cuda_arch_compat.py:19
↓ 3 callers
Method
_deactivate_upload
(self, upload: _Upload)
backend/worker/transport/server.py:2810
↓ 3 callers
Function
_decode
()
tests/test_asr_ffmpeg_decode.py:22
↓ 3 callers
Function
_decode_audio_16k_mono
Decode `audio_path` to a 16 kHz mono float32 waveform using VoiceStudio's *validated* ffmpeg, instead of whisperx.load_audio's bare ``"ffmpeg"`` P
backend/services/asr_backend.py:233
↓ 3 callers
Function
_decode_ref_audio
Decoded reference audio, or None when the input isn't valid base64. LLM agents frequently prepend a data URI (``data:audio/wav;base64,…``) wh
backend/mcp_server.py:33
↓ 3 callers
Function
_default_models_dir
huggingface_hub's default cache root, honoring XDG_CACHE_HOME on Linux (matches HF so GET reports the *true* default the backend would use).
backend/api/routers/settings.py:729
↓ 3 callers
Method
_degrade_for_vram
Downgrade the CUDA compute type (or fall to CPU) if free VRAM can't hold the model — preventing the un-catchable native OOM abort (#723).
backend/services/asr_backend.py:558
↓ 3 callers
Function
_detect_gpu
Best-effort detection of GPU vendor + driver + compute backend.
backend/api/routers/setup/wizard.py:84
↓ 3 callers
Function
_dialect_flags
Response fields describing whether the requested dialect was honored. Empty dict when no dialect was requested, so existing response shapes s
backend/api/routers/dub_translate.py:171
↓ 3 callers
Function
_download_endpoint
Optional HF endpoint override, per-call ``endpoint=`` rather than a process-wide HF_ENDPOINT mutation. Explicit configuration (FDL-10 mirror p
backend/api/routers/setup/download.py:94
↓ 3 callers
Function
_drain
(dc, job_id)
tests/test_dub_no_tts_load_for_asr.py:100
↓ 3 callers
Method
_drop
(self, endpoint: str)
backend/worker/inbound/service.py:660
↓ 3 callers
Function
_effective_models_dir
()
backend/api/routers/settings.py:736
↓ 3 callers
Function
_emit
Fan out to all registered listeners. Never raise — a bad listener shouldn't break a download.
backend/utils/hf_progress.py:97
↓ 3 callers
Function
_endpoint_of
(entry: str)
backend/worker/inbound/service.py:848
↓ 3 callers
Function
_ensure_durable_directory
Create a staging hierarchy and persist each new directory entry.
backend/worker/inbound/artifacts.py:1132
↓ 3 callers
Method
_ensure_loaded
(self)
backend/services/tts_backend.py:1057
↓ 3 callers
Method
_ensure_rec
(self)
backend/services/asr_backend.py:1775
↓ 3 callers
Function
_env
()
tests/test_published_image_path_is_pinned.py:25
↓ 3 callers
Function
_env_first
(names: tuple[str, ...])
backend/services/llm_providers.py:185
↓ 3 callers
Function
_env_flag
(name: str, default: bool = False)
backend/services/model_manager.py:1793
↓ 3 callers
Function
_env_float
Parse a float env override, rejecting negative and non-finite values. Shared by the preload-delay / timeout knobs: NaN would silently never f
backend/main.py:390
↓ 3 callers
Function
_fake_audioseal
Install a fake ``audioseal`` module whose load blocks ``load_s`` and records every invocation. The block is what makes a missing lock fail rel
tests/test_watermark_prefetch_coldstart.py:44
↓ 3 callers
Function
_fake_run_logged
A _run_logged stand-in that fabricates git/uv side effects on disk.
tests/test_sidecar_install.py:76
↓ 3 callers
Function
_fake_wheel
(version: str)
tests/test_media_tools.py:332
↓ 3 callers
Function
_fallbackHeaders
Auth context a non-desktop deployment may need (LAN-share PIN, ephemeral * admin session) — mirrors apiFetch's injection. We deliberately do NOT call
frontend/src/utils/backendCrash.ts:123
↓ 3 callers
Method
_fetch_inbound_artifact
Pull a staged result down from a node this control plane dialled. Returns the local path, or None — and None is not a silent loss: the
backend/worker/transport/server.py:2415
↓ 3 callers
Function
_file_size
(path: str)
backend/worker/inbound/artifacts.py:1077
↓ 3 callers
Function
_file_version
Fields that identify both a staged path and the bytes hashed from it.
backend/worker/inbound/connector.py:79
↓ 3 callers
Function
_finalize_generation
Shared tail of a successful generation: watermark → save WAV → history row (self-healing) → retention prune → event emit. Used verbatim by bo
backend/api/routers/generation.py:997
↓ 3 callers
Function
_fitted_segments_for
Fitted-timeline subtitle cues ({id, start, end}) for a Smart Fit track, or None. Same staleness gate as ``_video_retime_plan_for``.
backend/api/routers/dub_export.py:484
↓ 3 callers
Function
_force_host
(monkeypatch, family="cpu")
tests/test_generate_engine.py:161
↓ 3 callers
Function
_force_host
Pin detect_host_caps() to a specific host family so routing is deterministic regardless of the CI runner's actual hardware.
tests/backend/api/test_engines_route_shape.py:254
↓ 3 callers
Function
_force_notimplemented
(monkeypatch)
tests/test_subprocess_fallback.py:17
↓ 3 callers
Function
_format_srt_time
(seconds)
backend/api/routers/dub_export.py:1609
↓ 3 callers
Function
_forward_logits
LLM forward + audio head over target positions only. The scoring step consumes logits at the cond-target and uncond ranges (2*sum(t_len) of t
omnivoice/models/omnivoice_flashinfer.py:456
↓ 3 callers
Function
_frame_rms
RMS per non-overlapping frame; the ragged tail frame is dropped.
backend/services/onset_align.py:94
↓ 3 callers
Function
_frames_bytes
(raw: bytes)
tests/test_indextts_config_and_timeout_1611.py:198
↓ 3 callers
Function
_fsync_parent_directory
Make a preceding directory-entry replacement durable where supported.
backend/worker/agent.py:514
↓ 3 callers
Function
_fsync_parent_directory
Persist directory entry changes where the platform supports it.
backend/worker/task_store.py:149
↓ 3 callers
Function
_fsync_parent_directory
Make a preceding directory-entry change durable where supported.
backend/worker/tls.py:444
↓ 3 callers
Function
_fsync_parent_directory
Persist a rename/create where the platform supports directory fsync.
backend/worker/transport/server.py:152
↓ 3 callers
Function
_generate_timeout_s
Wall-clock budget for one generate, scaled to the request. Thin alias for the canonical helper, which moved to ``services.model_manager.gener
backend/api/routers/generation.py:717
↓ 3 callers
Function
_get
(repo_id: str, target: str = "local")
backend/utils/download_aggregator.py:189
↓ 3 callers
Method
_get
()
tests/test_api.py:519
↓ 3 callers
Function
_get_clone_prompt
Return a cached/precomputed ``VoiceClonePrompt`` for (ref_audio, ref_text, preprocess_prompt), or ``None`` to fall back to the inline ref path
backend/services/tts_backend.py:556
↓ 3 callers
Function
_get_semaphore
()
backend/services/ffmpeg_utils.py:132
↓ 3 callers
Function
_gumbel_sample
(logits: torch.Tensor, temperature: float)
omnivoice/models/omnivoice.py:1704
↓ 3 callers
Function
_has_dedicated_vram
Check if the current device has limited dedicated VRAM that needs offloading.
backend/services/model_manager.py:3146
↓ 3 callers
Function
_healed_summary
(repo_id="test/checkpoint")
tests/test_hf_cache_repair.py:340
↓ 3 callers
Function
_hf_mirror_state
The full GET /hf-mirror payload. Auto info comes from the CACHED race decision only — reading settings never probes the network.
backend/api/routers/settings.py:894
↓ 3 callers
Function
_hub_cache_roots
Candidate roots that directly contain ``models--*`` dirs. HF stores repos under ``$HF_HUB_CACHE`` (== ``$HF_HOME/hub`` by default). When only
backend/api/routers/setup/models.py:284
↓ 3 callers
Function
_in_subprocess
Run `code` in a clean interpreter — import side effects don't survive into it, so `sys.modules` assertions actually mean something.
tests/test_omnivoice_lazy_model_import.py:28
↓ 3 callers
Method
_init_queue
(self)
backend/core/tasks.py:29
↓ 3 callers
Function
_insert_profile
(pid, **cols)
tests/test_generate_ref_text_persist.py:87
↓ 3 callers
Function
_install_fake_argos
Register a fake `argostranslate` package that translates en→es to `[es]<text>` with a pre-installed package, so the argos branch runs offline.
tests/test_dub_translate.py:237
↓ 3 callers
Function
_install_fake_openai
Register a fake `openai.OpenAI` whose chat.completions.create returns `content` (or raises `raises`). Accepts the max_retries kwarg the code adds.
tests/test_dub_translate.py:327
↓ 3 callers
Function
_is_blank_audio
True if a render came back effectively silent / empty / non-finite. After ``normalize_audio``'s silence-floor guard a dead render stays at the
backend/api/routers/archetypes.py:228
↓ 3 callers
Function
_is_compute_type_error
(msg: str)
backend/services/asr_backend.py:204
↓ 3 callers
Function
_is_end_control
Accept the versioned JSON control frame and the legacy ``EOF`` frame.
backend/api/routers/capture_ws.py:92
↓ 3 callers
Function
_is_transient_download_error
True when a download failure is worth retrying (broken pipe / net drop). Reuses the single failure taxonomy (`VIDEO_DOWNLOAD_NETWORK`) rather tha
backend/services/dub_pipeline.py:856
↓ 3 callers
Method
_is_voice_design
Whether the loaded model builds a voice from a text description. Asks the model's own config — `tts_model_type`, the exact field mlx-
backend/services/tts_backend.py:1637
↓ 3 callers
Function
_iter_frontend_files
(suffixes)
tests/test_no_literal_borders.py:89
↓ 3 callers
Function
_join_url
(base_url: str, path: str)
backend/speech_client/__main__.py:33
↓ 3 callers
Function
_lazy_omnivoice
()
backend/services/model_manager.py:60
↓ 3 callers
Function
_live
()
tests/test_omnivoice_path.py:60
↓ 3 callers
Method
_live_pool
(self)
backend/services/model_manager.py:257
↓ 3 callers
Function
_load
()
backend/core/prefs.py:36
↓ 3 callers
Function
_load_locales
Return ({locale_name: flattened_keys}, [bad_files]).
tests/probe/judges/i18n.py:31
↓ 3 callers
Function
_load_store
()
backend/core/run_sentinel.py:301
↓ 3 callers
Function
_locate_uv
Find uv: bundled (Tauri-set OMNIVOICE_BUNDLED_UV) first, then PATH. Same resolution order as engines.indextts.bootstrap._locate_uv — the cano
backend/services/sidecar_install.py:203
↓ 3 callers
Function
_log_safe
Backward-compatible alias for the shared logging seam.
backend/services/model_manager.py:613
↓ 3 callers
Function
_make_coro
(value)
backend/tests/test_stream_stderr_sync_fallback.py:142
↓ 3 callers
Function
_make_epub
(n_chapters: int = 3)
backend/tests/test_longform_import_parse_failure.py:18
↓ 3 callers
Function
_make_fake_cache
Create the canonical HF cache layout for repo_id under tmp_path.
tests/test_hf_cache_fallback.py:14
↓ 3 callers
Function
_make_fake_engine
(engine_id="fake-bridge-engine")
tests/test_pyvideotrans_contract.py:39
↓ 3 callers
Function
_make_fake_engine
(engine_id="fake-chunk-engine")
tests/test_chunked_tts.py:188
↓ 3 callers
Function
_make_fake_engine
(engine_id="fake-lang-engine", gpu_compat=("cpu",))
tests/test_profile_language_propagation.py:29
← previous
next →
1,301–1,400 of 15,058, ranked by callers