vocal-api 0.3.8__tar.gz → 0.4.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {vocal_api-0.3.8 → vocal_api-0.4.0}/.gitignore +1 -2
- {vocal_api-0.3.8 → vocal_api-0.4.0}/MANIFEST.in +4 -4
- {vocal_api-0.3.8 → vocal_api-0.4.0}/PKG-INFO +7 -7
- {vocal_api-0.3.8 → vocal_api-0.4.0}/pyproject.toml +6 -6
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/__init__.py +9 -9
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/config.py +3 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/main.py +60 -9
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/models/model.py +24 -4
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/routes/__init__.py +2 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/routes/models.py +17 -15
- vocal_api-0.4.0/vocal_api/routes/realtime.py +745 -0
- vocal_api-0.4.0/vocal_api/routes/settings.py +185 -0
- vocal_api-0.4.0/vocal_api/routes/stream.py +300 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/routes/system.py +16 -16
- vocal_api-0.4.0/vocal_api/routes/transcription.py +139 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/routes/tts.py +13 -2
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/services/__init__.py +9 -9
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/services/model_service.py +6 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/services/transcription_service.py +50 -21
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/services/tts_service.py +113 -16
- vocal_api-0.4.0/vocal_api/webui/app.css +472 -0
- vocal_api-0.4.0/vocal_api/webui/app.html +96 -0
- vocal_api-0.4.0/vocal_api/webui/design.md +137 -0
- vocal_api-0.4.0/vocal_api/webui/index.html +570 -0
- vocal_api-0.4.0/vocal_api/webui/js/api.js +155 -0
- vocal_api-0.4.0/vocal_api/webui/js/app.js +168 -0
- vocal_api-0.4.0/vocal_api/webui/js/audio.js +289 -0
- vocal_api-0.4.0/vocal_api/webui/js/connection.js +128 -0
- vocal_api-0.4.0/vocal_api/webui/js/icons.js +50 -0
- vocal_api-0.4.0/vocal_api/webui/js/mic.js +132 -0
- vocal_api-0.4.0/vocal_api/webui/js/pcm-worklet.js +11 -0
- vocal_api-0.4.0/vocal_api/webui/js/store.js +81 -0
- vocal_api-0.4.0/vocal_api/webui/js/ui.js +190 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/agent.js +417 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/clone.js +293 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/dashboard.js +163 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/live.js +240 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/models.js +290 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/settings.js +189 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/stt.js +344 -0
- vocal_api-0.4.0/vocal_api/webui/js/views/tts.js +230 -0
- vocal_api-0.4.0/vocal_api/webui/og.svg +33 -0
- vocal_api-0.4.0/vocal_api/webui/theme.css +585 -0
- vocal_api-0.3.8/vocal_api/routes/realtime.py +0 -399
- vocal_api-0.3.8/vocal_api/routes/stream.py +0 -216
- vocal_api-0.3.8/vocal_api/routes/transcription.py +0 -83
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/dependencies.py +0 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/models/__init__.py +0 -0
- {vocal_api-0.3.8 → vocal_api-0.4.0}/vocal_api/models/transcription.py +0 -0
|
@@ -1,4 +1,4 @@
|
|
|
1
|
-
include ../../../README.md
|
|
2
|
-
include ../../../LICENSE
|
|
3
|
-
recursive-include vocal_api *.py
|
|
4
|
-
recursive-include vocal_api/models *.py
|
|
1
|
+
include ../../../README.md
|
|
2
|
+
include ../../../LICENSE
|
|
3
|
+
recursive-include vocal_api *.py
|
|
4
|
+
recursive-include vocal_api/models *.py
|
|
@@ -1,6 +1,6 @@
|
|
|
1
|
-
Metadata-Version: 2.
|
|
1
|
+
Metadata-Version: 2.5
|
|
2
2
|
Name: vocal-api
|
|
3
|
-
Version: 0.
|
|
3
|
+
Version: 0.4.0
|
|
4
4
|
Summary: FastAPI server for Vocal - Ollama-style Voice Model Management
|
|
5
5
|
Project-URL: Homepage, https://github.com/niradler/vocal
|
|
6
6
|
Project-URL: Documentation, https://github.com/niradler/vocal/tree/master/docs
|
|
@@ -19,8 +19,8 @@ Classifier: Programming Language :: Python :: 3.13
|
|
|
19
19
|
Classifier: Topic :: Internet :: WWW/HTTP :: HTTP Servers
|
|
20
20
|
Requires-Python: >=3.11
|
|
21
21
|
Requires-Dist: aiofiles>=25.1.0
|
|
22
|
-
Requires-Dist: fastapi>=0.
|
|
23
|
-
Requires-Dist: pydantic>=2.
|
|
24
|
-
Requires-Dist: python-multipart>=0.0.
|
|
25
|
-
Requires-Dist: uvicorn[standard]>=0.
|
|
26
|
-
Requires-Dist: vocal-core>=0.
|
|
22
|
+
Requires-Dist: fastapi>=0.163.2
|
|
23
|
+
Requires-Dist: pydantic>=2.14.3
|
|
24
|
+
Requires-Dist: python-multipart>=0.0.34
|
|
25
|
+
Requires-Dist: uvicorn[standard]>=0.54.0
|
|
26
|
+
Requires-Dist: vocal-core>=0.4.0
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
[project]
|
|
2
2
|
name = "vocal-api"
|
|
3
|
-
version = "0.
|
|
3
|
+
version = "0.4.0"
|
|
4
4
|
description = "FastAPI server for Vocal - Ollama-style Voice Model Management"
|
|
5
5
|
requires-python = ">=3.11"
|
|
6
6
|
license = { text = "SSPL-1.0" }
|
|
@@ -19,11 +19,11 @@ classifiers = [
|
|
|
19
19
|
"Topic :: Internet :: WWW/HTTP :: HTTP Servers",
|
|
20
20
|
]
|
|
21
21
|
dependencies = [
|
|
22
|
-
"fastapi>=0.
|
|
23
|
-
"uvicorn[standard]>=0.
|
|
24
|
-
"pydantic>=2.
|
|
25
|
-
"python-multipart>=0.0.
|
|
26
|
-
"vocal-core>=0.
|
|
22
|
+
"fastapi>=0.163.2",
|
|
23
|
+
"uvicorn[standard]>=0.54.0",
|
|
24
|
+
"pydantic>=2.14.3",
|
|
25
|
+
"python-multipart>=0.0.34",
|
|
26
|
+
"vocal-core>=0.4.0",
|
|
27
27
|
"aiofiles>=25.1.0",
|
|
28
28
|
]
|
|
29
29
|
|
|
@@ -1,9 +1,9 @@
|
|
|
1
|
-
from .config import settings
|
|
2
|
-
from .main import app
|
|
3
|
-
|
|
4
|
-
__version__ = settings.VERSION
|
|
5
|
-
|
|
6
|
-
__all__ = [
|
|
7
|
-
"app",
|
|
8
|
-
"settings",
|
|
9
|
-
]
|
|
1
|
+
from .config import settings
|
|
2
|
+
from .main import app
|
|
3
|
+
|
|
4
|
+
__version__ = settings.VERSION
|
|
5
|
+
|
|
6
|
+
__all__ = [
|
|
7
|
+
"app",
|
|
8
|
+
"settings",
|
|
9
|
+
]
|
|
@@ -16,6 +16,9 @@ class Settings(BaseSettings):
|
|
|
16
16
|
DEBUG: bool = False
|
|
17
17
|
|
|
18
18
|
CORS_ORIGINS: list[str] = ["*"]
|
|
19
|
+
# Allow browser requests from public origins (e.g. the GitHub Pages
|
|
20
|
+
# console) to this local server — Chrome's Private Network Access.
|
|
21
|
+
CORS_ALLOW_PRIVATE_NETWORK: bool = True
|
|
19
22
|
|
|
20
23
|
MAX_UPLOAD_SIZE: int = 25 * 1024 * 1024
|
|
21
24
|
|
|
@@ -17,22 +17,37 @@ if sys.platform == "win32":
|
|
|
17
17
|
del ctypes
|
|
18
18
|
|
|
19
19
|
from contextlib import asynccontextmanager
|
|
20
|
+
from pathlib import Path
|
|
20
21
|
|
|
21
|
-
from fastapi import FastAPI
|
|
22
|
+
from fastapi import FastAPI, Request
|
|
22
23
|
from fastapi.middleware.cors import CORSMiddleware
|
|
23
24
|
from fastapi.openapi.utils import get_openapi
|
|
25
|
+
from fastapi.responses import FileResponse, JSONResponse
|
|
26
|
+
from fastapi.staticfiles import StaticFiles
|
|
24
27
|
|
|
25
28
|
from vocal_core.config import vocal_settings
|
|
26
29
|
from vocal_core.logging import setup_logging
|
|
27
30
|
|
|
28
31
|
from .config import settings
|
|
29
32
|
from .dependencies import get_transcription_service, get_tts_service
|
|
30
|
-
from .routes import
|
|
33
|
+
from .routes import (
|
|
34
|
+
models_router,
|
|
35
|
+
realtime_router,
|
|
36
|
+
settings_router,
|
|
37
|
+
stream_router,
|
|
38
|
+
system_router,
|
|
39
|
+
transcription_router,
|
|
40
|
+
tts_router,
|
|
41
|
+
)
|
|
42
|
+
from .routes.settings import load_persisted_settings
|
|
43
|
+
|
|
44
|
+
WEBUI_DIR = Path(__file__).parent / "webui"
|
|
31
45
|
|
|
32
46
|
|
|
33
47
|
@asynccontextmanager
|
|
34
48
|
async def lifespan(app: FastAPI):
|
|
35
49
|
"""Initialize services and start background tasks"""
|
|
50
|
+
load_persisted_settings()
|
|
36
51
|
setup_logging(level=vocal_settings.LOG_LEVEL, fmt=vocal_settings.LOG_FORMAT)
|
|
37
52
|
|
|
38
53
|
try:
|
|
@@ -68,24 +83,54 @@ app.add_middleware(
|
|
|
68
83
|
allow_credentials=True,
|
|
69
84
|
allow_methods=["*"],
|
|
70
85
|
allow_headers=["*"],
|
|
86
|
+
# Lets a publicly hosted console (GitHub Pages) reach this local server
|
|
87
|
+
# from the browser (Chrome Private Network Access preflight).
|
|
88
|
+
allow_private_network=settings.CORS_ALLOW_PRIVATE_NETWORK,
|
|
71
89
|
)
|
|
72
90
|
|
|
73
91
|
app.include_router(transcription_router)
|
|
74
92
|
app.include_router(models_router)
|
|
75
93
|
app.include_router(tts_router)
|
|
76
94
|
app.include_router(system_router)
|
|
95
|
+
app.include_router(settings_router)
|
|
77
96
|
app.include_router(stream_router)
|
|
78
97
|
app.include_router(realtime_router)
|
|
79
98
|
|
|
99
|
+
# Web UI static assets (landing page + console). Served under /static so the
|
|
100
|
+
# HTML entry points below can reference shared CSS/JS without a build step.
|
|
101
|
+
app.mount("/static", StaticFiles(directory=WEBUI_DIR), name="static")
|
|
102
|
+
|
|
80
103
|
|
|
81
104
|
@app.get("/", tags=["health"])
|
|
82
|
-
async def root():
|
|
83
|
-
"""API health check
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
87
|
-
|
|
88
|
-
|
|
105
|
+
async def root(request: Request):
|
|
106
|
+
"""API health check for clients; landing page for browsers.
|
|
107
|
+
|
|
108
|
+
Browsers send ``Accept: text/html`` and get the landing page. API clients
|
|
109
|
+
(SDK, curl, health checks) send ``*/*`` or ``application/json`` and keep the
|
|
110
|
+
original JSON contract.
|
|
111
|
+
"""
|
|
112
|
+
accept = request.headers.get("accept", "")
|
|
113
|
+
if "text/html" in accept:
|
|
114
|
+
return FileResponse(WEBUI_DIR / "index.html", media_type="text/html", headers={"Vary": "Accept"})
|
|
115
|
+
return JSONResponse(
|
|
116
|
+
{
|
|
117
|
+
"status": "ok",
|
|
118
|
+
"message": "Vocal API - Ollama-style voice model management",
|
|
119
|
+
"version": settings.VERSION,
|
|
120
|
+
}
|
|
121
|
+
)
|
|
122
|
+
|
|
123
|
+
|
|
124
|
+
@app.get("/welcome", include_in_schema=False)
|
|
125
|
+
async def welcome():
|
|
126
|
+
"""Landing page (explicit route)."""
|
|
127
|
+
return FileResponse(WEBUI_DIR / "index.html", media_type="text/html")
|
|
128
|
+
|
|
129
|
+
|
|
130
|
+
@app.get("/ui", include_in_schema=False)
|
|
131
|
+
async def console():
|
|
132
|
+
"""Web console — manage models, run speech pipelines, configure Vocal."""
|
|
133
|
+
return FileResponse(WEBUI_DIR / "app.html", media_type="text/html")
|
|
89
134
|
|
|
90
135
|
|
|
91
136
|
@app.get("/health", tags=["health"])
|
|
@@ -113,3 +158,9 @@ def custom_openapi():
|
|
|
113
158
|
|
|
114
159
|
|
|
115
160
|
app.openapi = custom_openapi
|
|
161
|
+
|
|
162
|
+
# Fallback static mount: serves webui assets at root-level paths (/theme.css,
|
|
163
|
+
# /js/app.js, ...) so the HTML entry points can use relative URLs. The same
|
|
164
|
+
# files then also work under a GitHub Pages subpath. Mounted LAST so every API
|
|
165
|
+
# route above always wins.
|
|
166
|
+
app.mount("/", StaticFiles(directory=WEBUI_DIR), name="webui")
|
|
@@ -25,13 +25,12 @@ class ModelBackend(str, Enum):
|
|
|
25
25
|
FASTER_QWEN3_TTS = "faster_qwen3_tts"
|
|
26
26
|
PIPER = "piper"
|
|
27
27
|
CHATTERBOX = "chatterbox"
|
|
28
|
-
XTTS = "xtts"
|
|
29
|
-
FISH_SPEECH = "fish_speech"
|
|
30
|
-
ORPHEUS = "orpheus"
|
|
31
|
-
DIA = "dia"
|
|
32
28
|
VOXTRAL_STT = "voxtral_stt"
|
|
33
29
|
VOXTRAL_TTS = "voxtral_tts"
|
|
34
30
|
OMNIVOICE = "omnivoice"
|
|
31
|
+
SUPERTONIC = "supertonic"
|
|
32
|
+
VOXCPM = "voxcpm"
|
|
33
|
+
PHONON = "phonon"
|
|
35
34
|
CUSTOM = "custom"
|
|
36
35
|
|
|
37
36
|
|
|
@@ -50,6 +49,21 @@ class ModelTask(str, Enum):
|
|
|
50
49
|
TTS = "tts"
|
|
51
50
|
|
|
52
51
|
|
|
52
|
+
class ModelRuntime(str, Enum):
|
|
53
|
+
"""How a model is executed"""
|
|
54
|
+
|
|
55
|
+
IN_PROCESS = "in_process"
|
|
56
|
+
EXTERNAL = "external"
|
|
57
|
+
|
|
58
|
+
|
|
59
|
+
class ModelMaintenanceStatus(str, Enum):
|
|
60
|
+
"""Upstream maintenance state of the model"""
|
|
61
|
+
|
|
62
|
+
ACTIVE = "active"
|
|
63
|
+
RETIRED = "retired"
|
|
64
|
+
DEPRECATED = "deprecated"
|
|
65
|
+
|
|
66
|
+
|
|
53
67
|
class ModelInfo(BaseModel):
|
|
54
68
|
"""Model information schema"""
|
|
55
69
|
|
|
@@ -67,6 +81,10 @@ class ModelInfo(BaseModel):
|
|
|
67
81
|
status: ModelStatus = Field(description="Current model status")
|
|
68
82
|
source_url: HttpUrl | None = None
|
|
69
83
|
license: str | None = None
|
|
84
|
+
commercial_use: bool | None = Field(default=None, description="Whether the model weights license permits commercial use")
|
|
85
|
+
runtime: ModelRuntime = Field(default=ModelRuntime.IN_PROCESS, description="'in_process' when Vocal runs the model itself, 'external' when it needs an out-of-process engine")
|
|
86
|
+
runtime_note: str | None = Field(default=None, description="Engine/serving requirement outside Vocal's bundled adapters, if any")
|
|
87
|
+
maintenance_status: ModelMaintenanceStatus = Field(default=ModelMaintenanceStatus.ACTIVE, description="Upstream maintenance state (active/retired/deprecated)")
|
|
70
88
|
recommended_vram: str | None = None
|
|
71
89
|
task: ModelTask = Field(description="Task type")
|
|
72
90
|
local_path: str | None = None
|
|
@@ -79,6 +97,7 @@ class ModelInfo(BaseModel):
|
|
|
79
97
|
sha: str | None = Field(default=None, description="Git commit SHA")
|
|
80
98
|
files: list[dict] | None = Field(default=None, description="List of model files")
|
|
81
99
|
supports_streaming: bool = Field(default=False, description="Whether the model supports true streaming output")
|
|
100
|
+
supports_live_streaming: bool = Field(default=False, description="Whether the model supports true chunk-by-chunk live audio streaming")
|
|
82
101
|
supports_voice_list: bool = Field(default=False, description="Whether the model exposes selectable voices")
|
|
83
102
|
supports_voice_clone: bool = Field(default=False, description="Whether the model supports reference-audio voice cloning")
|
|
84
103
|
supports_voice_design: bool = Field(default=False, description="Whether the model supports prompt/instruction-based voice design")
|
|
@@ -87,6 +106,7 @@ class ModelInfo(BaseModel):
|
|
|
87
106
|
clone_mode: str | None = Field(default=None, description="How /v1/audio/clone works for this model")
|
|
88
107
|
reference_audio_min_seconds: float | None = Field(default=None, description="Recommended minimum reference audio length for cloning")
|
|
89
108
|
reference_audio_max_seconds: float | None = Field(default=None, description="Recommended maximum reference audio length for cloning")
|
|
109
|
+
output_formats: list[str] = Field(default_factory=list, description="Audio formats /v1/audio/speech accepts for this model (empty for STT)")
|
|
90
110
|
|
|
91
111
|
|
|
92
112
|
class ModelListResponse(BaseModel):
|
|
@@ -1,5 +1,6 @@
|
|
|
1
1
|
from .models import router as models_router
|
|
2
2
|
from .realtime import router as realtime_router
|
|
3
|
+
from .settings import router as settings_router
|
|
3
4
|
from .stream import router as stream_router
|
|
4
5
|
from .system import router as system_router
|
|
5
6
|
from .transcription import router as transcription_router
|
|
@@ -10,6 +11,7 @@ __all__ = [
|
|
|
10
11
|
"models_router",
|
|
11
12
|
"tts_router",
|
|
12
13
|
"system_router",
|
|
14
|
+
"settings_router",
|
|
13
15
|
"stream_router",
|
|
14
16
|
"realtime_router",
|
|
15
17
|
]
|
|
@@ -46,21 +46,6 @@ async def list_catalog(
|
|
|
46
46
|
return ModelListResponse(models=models, total=len(models))
|
|
47
47
|
|
|
48
48
|
|
|
49
|
-
@router.get(
|
|
50
|
-
"/{model_id:path}",
|
|
51
|
-
response_model=ModelInfo,
|
|
52
|
-
response_model_exclude_none=True,
|
|
53
|
-
summary="Get model info",
|
|
54
|
-
description="Get detailed information about a specific model",
|
|
55
|
-
)
|
|
56
|
-
async def get_model(model_id: str, service: ModelService = Depends(get_model_service)) -> ModelInfo:
|
|
57
|
-
"""Get detailed model information"""
|
|
58
|
-
model = await service.get_model(model_id)
|
|
59
|
-
if not model:
|
|
60
|
-
raise HTTPException(404, f"Model {model_id} not found")
|
|
61
|
-
return model
|
|
62
|
-
|
|
63
|
-
|
|
64
49
|
@router.post(
|
|
65
50
|
"/{model_id:path}/download",
|
|
66
51
|
response_model=ModelDownloadProgress,
|
|
@@ -128,6 +113,23 @@ async def get_download_status(model_id: str, service: ModelService = Depends(get
|
|
|
128
113
|
return status
|
|
129
114
|
|
|
130
115
|
|
|
116
|
+
# NOTE: the greedy {model_id:path} routes must be declared AFTER the specific
|
|
117
|
+
# /download and /download/status suffixes, or they swallow those paths whole.
|
|
118
|
+
@router.get(
|
|
119
|
+
"/{model_id:path}",
|
|
120
|
+
response_model=ModelInfo,
|
|
121
|
+
response_model_exclude_none=True,
|
|
122
|
+
summary="Get model info",
|
|
123
|
+
description="Get detailed information about a specific model",
|
|
124
|
+
)
|
|
125
|
+
async def get_model(model_id: str, service: ModelService = Depends(get_model_service)) -> ModelInfo:
|
|
126
|
+
"""Get detailed model information"""
|
|
127
|
+
model = await service.get_model(model_id)
|
|
128
|
+
if not model:
|
|
129
|
+
raise HTTPException(404, f"Model {model_id} not found")
|
|
130
|
+
return model
|
|
131
|
+
|
|
132
|
+
|
|
131
133
|
@router.delete("/{model_id:path}", summary="Delete model", description="Remove a downloaded model")
|
|
132
134
|
async def delete_model(model_id: str, service: ModelService = Depends(get_model_service)):
|
|
133
135
|
"""Delete a downloaded model"""
|