Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -3,7 +3,8 @@ import os
|
|
| 3 |
import tempfile
|
| 4 |
import uuid
|
| 5 |
from pathlib import Path
|
| 6 |
-
from
|
|
|
|
| 7 |
|
| 8 |
import requests
|
| 9 |
import torch
|
|
@@ -24,6 +25,8 @@ MAX_TEXT_LENGTH = 1000
|
|
| 24 |
DEFAULT_LANGUAGE = "en"
|
| 25 |
|
| 26 |
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
|
|
|
|
|
|
|
| 27 |
|
| 28 |
# Set token in environment before importing
|
| 29 |
if HF_TOKEN:
|
|
@@ -148,10 +151,20 @@ def _preprocess_audio_wav(path: str, target_sr: int = 24000, target_peak: float
|
|
| 148 |
return path
|
| 149 |
|
| 150 |
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
| 154 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 155 |
|
| 156 |
|
| 157 |
def _cleanup_files(*files: str):
|
|
@@ -164,59 +177,106 @@ def _cleanup_files(*files: str):
|
|
| 164 |
pass # Ignore cleanup errors
|
| 165 |
|
| 166 |
|
| 167 |
-
|
| 168 |
-
def generate(
|
| 169 |
-
payload: GenerateRequest = Body(...),
|
| 170 |
-
background_tasks: BackgroundTasks = BackgroundTasks(),
|
| 171 |
-
x_api_key: Optional[str] = Header(default=None),
|
| 172 |
-
):
|
| 173 |
-
_require_api_key(x_api_key)
|
| 174 |
-
|
| 175 |
speaker_file = None
|
| 176 |
output_file = None
|
| 177 |
-
|
| 178 |
try:
|
| 179 |
-
speaker_file = _temp_speaker_file(payload
|
| 180 |
speaker_file = _preprocess_audio_wav(speaker_file)
|
| 181 |
output_file = os.path.join(tempfile.gettempdir(), f"indextts2-{uuid.uuid4()}.wav")
|
| 182 |
|
| 183 |
-
# IndexTTS2 inference
|
| 184 |
-
# Note: language parameter is kept for API compatibility but IndexTTS2
|
| 185 |
-
# handles multilingual automatically (supports English, Turkish, Chinese, etc.)
|
| 186 |
tts_model.infer(
|
| 187 |
spk_audio_prompt=speaker_file,
|
| 188 |
-
text=payload
|
| 189 |
output_path=output_file,
|
| 190 |
-
use_random=False,
|
| 191 |
verbose=False,
|
| 192 |
)
|
| 193 |
|
| 194 |
-
# Light post-process to avoid end-of-file artifacts
|
| 195 |
output_file = _preprocess_audio_wav(output_file)
|
| 196 |
|
| 197 |
-
# Verify the output file was created
|
| 198 |
if not Path(output_file).exists():
|
| 199 |
raise RuntimeError(f"TTS generation failed: output file was not created at {output_file}")
|
| 200 |
|
| 201 |
-
|
| 202 |
-
|
|
|
|
|
|
|
|
|
|
| 203 |
|
| 204 |
-
return FileResponse(output_file, media_type="audio/wav", filename="output.wav")
|
| 205 |
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
|
| 209 |
-
|
| 210 |
-
|
| 211 |
-
|
| 212 |
-
|
| 213 |
-
|
| 214 |
-
|
| 215 |
-
|
| 216 |
-
|
| 217 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
|
| 219 |
|
| 220 |
@app.get("/")
|
| 221 |
def root():
|
| 222 |
-
return {
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
import tempfile
|
| 4 |
import uuid
|
| 5 |
from pathlib import Path
|
| 6 |
+
from threading import Lock
|
| 7 |
+
from typing import Dict, Optional
|
| 8 |
|
| 9 |
import requests
|
| 10 |
import torch
|
|
|
|
| 25 |
DEFAULT_LANGUAGE = "en"
|
| 26 |
|
| 27 |
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
|
| 28 |
+
JOBS: Dict[str, Dict[str, str]] = {}
|
| 29 |
+
JOB_LOCK = Lock()
|
| 30 |
|
| 31 |
# Set token in environment before importing
|
| 32 |
if HF_TOKEN:
|
|
|
|
| 151 |
return path
|
| 152 |
|
| 153 |
|
| 154 |
+
def _set_job(job_id: str, **kwargs):
|
| 155 |
+
with JOB_LOCK:
|
| 156 |
+
JOBS[job_id] = {**JOBS.get(job_id, {}), **kwargs}
|
| 157 |
+
|
| 158 |
+
|
| 159 |
+
def _get_job(job_id: str) -> Optional[Dict[str, str]]:
|
| 160 |
+
with JOB_LOCK:
|
| 161 |
+
data = JOBS.get(job_id)
|
| 162 |
+
return dict(data) if data else None
|
| 163 |
+
|
| 164 |
+
|
| 165 |
+
def _pop_job(job_id: str) -> Optional[Dict[str, str]]:
|
| 166 |
+
with JOB_LOCK:
|
| 167 |
+
return JOBS.pop(job_id, None)
|
| 168 |
|
| 169 |
|
| 170 |
def _cleanup_files(*files: str):
|
|
|
|
| 177 |
pass # Ignore cleanup errors
|
| 178 |
|
| 179 |
|
| 180 |
+
def _run_generate_job(job_id: str, payload: Dict[str, str]):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 181 |
speaker_file = None
|
| 182 |
output_file = None
|
| 183 |
+
_set_job(job_id, status="processing")
|
| 184 |
try:
|
| 185 |
+
speaker_file = _temp_speaker_file(payload["speaker_wav"])
|
| 186 |
speaker_file = _preprocess_audio_wav(speaker_file)
|
| 187 |
output_file = os.path.join(tempfile.gettempdir(), f"indextts2-{uuid.uuid4()}.wav")
|
| 188 |
|
|
|
|
|
|
|
|
|
|
| 189 |
tts_model.infer(
|
| 190 |
spk_audio_prompt=speaker_file,
|
| 191 |
+
text=payload["text"],
|
| 192 |
output_path=output_file,
|
| 193 |
+
use_random=False,
|
| 194 |
verbose=False,
|
| 195 |
)
|
| 196 |
|
|
|
|
| 197 |
output_file = _preprocess_audio_wav(output_file)
|
| 198 |
|
|
|
|
| 199 |
if not Path(output_file).exists():
|
| 200 |
raise RuntimeError(f"TTS generation failed: output file was not created at {output_file}")
|
| 201 |
|
| 202 |
+
_cleanup_files(speaker_file)
|
| 203 |
+
_set_job(job_id, status="completed", output_file=output_file)
|
| 204 |
+
except Exception as exc:
|
| 205 |
+
_cleanup_files(speaker_file, output_file)
|
| 206 |
+
_set_job(job_id, status="error", error=str(exc))
|
| 207 |
|
|
|
|
| 208 |
|
| 209 |
+
@app.post("/health")
|
| 210 |
+
def health(x_api_key: Optional[str] = Header(default=None)):
|
| 211 |
+
_require_api_key(x_api_key)
|
| 212 |
+
return {"status": "ok", "model": "indextts2", "device": DEVICE}
|
| 213 |
+
|
| 214 |
+
|
| 215 |
+
@app.post("/generate")
|
| 216 |
+
def generate(
|
| 217 |
+
payload: GenerateRequest = Body(...),
|
| 218 |
+
background_tasks: BackgroundTasks = BackgroundTasks(),
|
| 219 |
+
x_api_key: Optional[str] = Header(default=None),
|
| 220 |
+
):
|
| 221 |
+
_require_api_key(x_api_key)
|
| 222 |
+
job_id = str(uuid.uuid4())
|
| 223 |
+
_set_job(job_id, status="queued")
|
| 224 |
+
|
| 225 |
+
# Offload the long-running synthesis so the HTTP request stays fast (<100s)
|
| 226 |
+
background_tasks.add_task(_run_generate_job, job_id, payload.dict())
|
| 227 |
+
|
| 228 |
+
return JSONResponse(
|
| 229 |
+
status_code=202,
|
| 230 |
+
content={
|
| 231 |
+
"job_id": job_id,
|
| 232 |
+
"status": "queued",
|
| 233 |
+
"status_url": f"/status/{job_id}",
|
| 234 |
+
"result_url": f"/result/{job_id}",
|
| 235 |
+
},
|
| 236 |
+
)
|
| 237 |
+
|
| 238 |
+
|
| 239 |
+
@app.get("/status/{job_id}")
|
| 240 |
+
def job_status(job_id: str, x_api_key: Optional[str] = Header(default=None)):
|
| 241 |
+
_require_api_key(x_api_key)
|
| 242 |
+
job = _get_job(job_id)
|
| 243 |
+
if not job:
|
| 244 |
+
raise HTTPException(status_code=404, detail="Job not found")
|
| 245 |
+
payload: Dict[str, str] = {"job_id": job_id, "status": job.get("status", "unknown")}
|
| 246 |
+
if "error" in job:
|
| 247 |
+
payload["error"] = job["error"]
|
| 248 |
+
return payload
|
| 249 |
+
|
| 250 |
+
|
| 251 |
+
@app.get("/result/{job_id}")
|
| 252 |
+
def job_result(
|
| 253 |
+
job_id: str,
|
| 254 |
+
background_tasks: BackgroundTasks = BackgroundTasks(),
|
| 255 |
+
x_api_key: Optional[str] = Header(default=None),
|
| 256 |
+
):
|
| 257 |
+
_require_api_key(x_api_key)
|
| 258 |
+
job = _get_job(job_id)
|
| 259 |
+
if not job:
|
| 260 |
+
raise HTTPException(status_code=404, detail="Job not found")
|
| 261 |
+
status = job.get("status")
|
| 262 |
+
if status != "completed":
|
| 263 |
+
raise HTTPException(status_code=409, detail=f"Job not ready (status={status})")
|
| 264 |
+
|
| 265 |
+
output_file = job.get("output_file")
|
| 266 |
+
if not output_file or not Path(output_file).exists():
|
| 267 |
+
_pop_job(job_id)
|
| 268 |
+
raise HTTPException(status_code=410, detail="Result expired or missing")
|
| 269 |
+
|
| 270 |
+
# Remove job from memory and cleanup output after sending
|
| 271 |
+
_pop_job(job_id)
|
| 272 |
+
background_tasks.add_task(_cleanup_files, output_file)
|
| 273 |
+
|
| 274 |
+
return FileResponse(output_file, media_type="audio/wav", filename="output.wav")
|
| 275 |
|
| 276 |
|
| 277 |
@app.get("/")
|
| 278 |
def root():
|
| 279 |
+
return {
|
| 280 |
+
"name": "indextts2-api",
|
| 281 |
+
"endpoints": ["/health", "/generate", "/status/{job_id}", "/result/{job_id}"],
|
| 282 |
+
}
|