Spaces:
Running
on
Zero
Running
on
Zero
impl ACTIVE_STREAM_FLAG
Browse files- app.py +11 -13
- app/canary_speech_engine.py +1 -1
- app/session_utils.py +32 -13
- app/streaming_audio_processor.py +15 -10
- app/utils.py +84 -80
app.py
CHANGED
|
@@ -28,6 +28,7 @@ from app.session_utils import (
|
|
| 28 |
register_session_hash_code,
|
| 29 |
reset_all_active_session_hash_code,
|
| 30 |
get_active_task_flag_file,
|
|
|
|
| 31 |
|
| 32 |
)
|
| 33 |
|
|
@@ -79,8 +80,7 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 79 |
gr.Timer(3.0).tick(fn=get_active_session_hash_code, outputs=sessions_table)
|
| 80 |
|
| 81 |
demo.load(fn=on_load, inputs=None, outputs=[session_hash_code, session_hash_code_box])
|
| 82 |
-
demo.unload(on_unload)
|
| 83 |
-
stop_streaming_flags = gr.State(value={"stop": False})
|
| 84 |
active_filepath = gr.State(value=next(iter(EXAMPLE_CONFIGS)))
|
| 85 |
|
| 86 |
with gr.Walkthrough(selected=0) as walkthrough:
|
|
@@ -144,7 +144,7 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 144 |
|
| 145 |
webrtc_stream.stream(
|
| 146 |
fn=read_and_stream_audio,
|
| 147 |
-
inputs=[active_filepath, session_hash_code,
|
| 148 |
outputs=[webrtc_stream],
|
| 149 |
trigger=start_stream_button.click,
|
| 150 |
concurrency_id="audio_stream",
|
|
@@ -267,7 +267,6 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 267 |
with gr.Row():
|
| 268 |
gr.Markdown("##### Transcription / Translation Result")
|
| 269 |
with gr.Row():
|
| 270 |
-
|
| 271 |
task_output = gr.Textbox(
|
| 272 |
label="Transcription / Translation Result",
|
| 273 |
show_label=False,
|
|
@@ -286,14 +285,11 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 286 |
|
| 287 |
stop_stream_button.click(
|
| 288 |
fn=stop_streaming,
|
| 289 |
-
inputs=[session_hash_code
|
| 290 |
-
outputs=[stop_streaming_flags],
|
| 291 |
)
|
| 292 |
|
| 293 |
def stop_task_fn(session_hash_code):
|
| 294 |
-
|
| 295 |
-
if os.path.exists(transcribe_active):
|
| 296 |
-
os.remove(transcribe_active)
|
| 297 |
yield "Task stopped by user."
|
| 298 |
|
| 299 |
|
|
@@ -308,7 +304,7 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 308 |
streaming_policy, alignatt_thr, waitk_lagging,
|
| 309 |
exclude_sink_frames, xatt_scores_layer, hallucinations_detector]
|
| 310 |
def start_transcription(
|
| 311 |
-
session_hash_code,
|
| 312 |
task_type, lang_source, lang_target,
|
| 313 |
chunk_secs, left_context_secs, right_context_secs,
|
| 314 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
@@ -338,7 +334,7 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 338 |
"""Stream transcription or translation results in real time."""
|
| 339 |
accumulated = ""
|
| 340 |
# Boucle sur le générateur de `task2()`
|
| 341 |
-
for result, status, current_chunk in
|
| 342 |
session_hash_code,
|
| 343 |
task_type, lang_source, lang_target,
|
| 344 |
chunk_secs, left_context_secs, right_context_secs,
|
|
@@ -348,7 +344,9 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 348 |
if status == "success":
|
| 349 |
yield accumulated + result, gr.update(visible=True,value=current_chunk , elem_classes=["info"]), gr.update(visible=False), gr.update(visible=True)
|
| 350 |
accumulated += result
|
| 351 |
-
elif status in ["
|
|
|
|
|
|
|
| 352 |
yield accumulated, gr.update(visible=True,value=result , elem_classes=[status]), gr.update(visible=True), gr.update(visible=False)
|
| 353 |
|
| 354 |
|
|
@@ -368,7 +366,7 @@ with gr.Blocks(theme=theme, css=css_style) as demo:
|
|
| 368 |
# start_task_button.click(
|
| 369 |
# fn=start_task,
|
| 370 |
# inputs=[
|
| 371 |
-
# session_hash_code,
|
| 372 |
# task_type, lang_source, lang_target,
|
| 373 |
# chunk_secs, left_context_secs, right_context_secs,
|
| 374 |
# streaming_policy, alignatt_thr, waitk_lagging,
|
|
|
|
| 28 |
register_session_hash_code,
|
| 29 |
reset_all_active_session_hash_code,
|
| 30 |
get_active_task_flag_file,
|
| 31 |
+
remove_active_task_flag_file
|
| 32 |
|
| 33 |
)
|
| 34 |
|
|
|
|
| 80 |
gr.Timer(3.0).tick(fn=get_active_session_hash_code, outputs=sessions_table)
|
| 81 |
|
| 82 |
demo.load(fn=on_load, inputs=None, outputs=[session_hash_code, session_hash_code_box])
|
| 83 |
+
demo.unload(fn=on_unload)
|
|
|
|
| 84 |
active_filepath = gr.State(value=next(iter(EXAMPLE_CONFIGS)))
|
| 85 |
|
| 86 |
with gr.Walkthrough(selected=0) as walkthrough:
|
|
|
|
| 144 |
|
| 145 |
webrtc_stream.stream(
|
| 146 |
fn=read_and_stream_audio,
|
| 147 |
+
inputs=[active_filepath, session_hash_code,gr.State(READ_SIZE)],
|
| 148 |
outputs=[webrtc_stream],
|
| 149 |
trigger=start_stream_button.click,
|
| 150 |
concurrency_id="audio_stream",
|
|
|
|
| 267 |
with gr.Row():
|
| 268 |
gr.Markdown("##### Transcription / Translation Result")
|
| 269 |
with gr.Row():
|
|
|
|
| 270 |
task_output = gr.Textbox(
|
| 271 |
label="Transcription / Translation Result",
|
| 272 |
show_label=False,
|
|
|
|
| 285 |
|
| 286 |
stop_stream_button.click(
|
| 287 |
fn=stop_streaming,
|
| 288 |
+
inputs=[session_hash_code],
|
|
|
|
| 289 |
)
|
| 290 |
|
| 291 |
def stop_task_fn(session_hash_code):
|
| 292 |
+
remove_active_task_flag_file(session_hash_code)
|
|
|
|
|
|
|
| 293 |
yield "Task stopped by user."
|
| 294 |
|
| 295 |
|
|
|
|
| 304 |
streaming_policy, alignatt_thr, waitk_lagging,
|
| 305 |
exclude_sink_frames, xatt_scores_layer, hallucinations_detector]
|
| 306 |
def start_transcription(
|
| 307 |
+
session_hash_code,
|
| 308 |
task_type, lang_source, lang_target,
|
| 309 |
chunk_secs, left_context_secs, right_context_secs,
|
| 310 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
|
|
| 334 |
"""Stream transcription or translation results in real time."""
|
| 335 |
accumulated = ""
|
| 336 |
# Boucle sur le générateur de `task2()`
|
| 337 |
+
for result, status, current_chunk in task(
|
| 338 |
session_hash_code,
|
| 339 |
task_type, lang_source, lang_target,
|
| 340 |
chunk_secs, left_context_secs, right_context_secs,
|
|
|
|
| 344 |
if status == "success":
|
| 345 |
yield accumulated + result, gr.update(visible=True,value=current_chunk , elem_classes=["info"]), gr.update(visible=False), gr.update(visible=True)
|
| 346 |
accumulated += result
|
| 347 |
+
elif status in ["warning","info" ]:
|
| 348 |
+
yield accumulated, gr.update(visible=True,value=result , elem_classes=[status]), gr.update(visible=False), gr.update(visible=True)
|
| 349 |
+
elif status in ["error", "done"]:
|
| 350 |
yield accumulated, gr.update(visible=True,value=result , elem_classes=[status]), gr.update(visible=True), gr.update(visible=False)
|
| 351 |
|
| 352 |
|
|
|
|
| 366 |
# start_task_button.click(
|
| 367 |
# fn=start_task,
|
| 368 |
# inputs=[
|
| 369 |
+
# session_hash_code,
|
| 370 |
# task_type, lang_source, lang_target,
|
| 371 |
# chunk_secs, left_context_secs, right_context_secs,
|
| 372 |
# streaming_policy, alignatt_thr, waitk_lagging,
|
app/canary_speech_engine.py
CHANGED
|
@@ -373,7 +373,7 @@ class CanarySpeechEngine(IStreamingSpeechEngine):
|
|
| 373 |
# logging.info(f"--- transcribe_chunk: took {duration_ms:.2f} ms ---")
|
| 374 |
|
| 375 |
# Return both the full segment transcription and the new diff
|
| 376 |
-
|
| 377 |
|
| 378 |
def finalize_segment(self):
|
| 379 |
"""
|
|
|
|
| 373 |
# logging.info(f"--- transcribe_chunk: took {duration_ms:.2f} ms ---")
|
| 374 |
|
| 375 |
# Return both the full segment transcription and the new diff
|
| 376 |
+
yield current_transcription, new_text
|
| 377 |
|
| 378 |
def finalize_segment(self):
|
| 379 |
"""
|
app/session_utils.py
CHANGED
|
@@ -9,6 +9,7 @@ import gradio as gr
|
|
| 9 |
|
| 10 |
TMP_DIR = os.getenv("TMP_DIR", "/tmp/canary_aed_streaming")
|
| 11 |
ACTIVE_SESSIONS_HASH_FILE = os.path.join(TMP_DIR, "active_session_hash_code.json")
|
|
|
|
| 12 |
ACTIVE_TASK_FLAG="task_active_"
|
| 13 |
NAME_FOLDER_CHUNKS="chunks_"
|
| 14 |
|
|
@@ -58,17 +59,17 @@ def on_load(request: gr.Request):
|
|
| 58 |
# ---------------------------
|
| 59 |
def on_unload(request: gr.Request):
|
| 60 |
"""Called when the visitor closes or refreshes the app."""
|
| 61 |
-
|
| 62 |
sessions = _read_session_hash_code()
|
| 63 |
|
| 64 |
-
if
|
| 65 |
-
sessions.pop(
|
| 66 |
_write_session_hash_code(sessions)
|
| 67 |
-
remove_session_hash_code_data(
|
| 68 |
-
unregister_session_hash_code_hash(
|
| 69 |
-
logging.info(f"[{
|
| 70 |
else:
|
| 71 |
-
logging.info(f"[{
|
| 72 |
|
| 73 |
def ensure_tmp_dir():
|
| 74 |
"""Ensures the base temporary directory exists."""
|
|
@@ -92,7 +93,8 @@ def reset_all_active_session_hash_code():
|
|
| 92 |
# --- Clean all flag files (stream + transcribe) ---
|
| 93 |
for f in os.listdir(TMP_DIR):
|
| 94 |
if (
|
| 95 |
-
f.startswith(f"{ACTIVE_TASK_FLAG}")
|
|
|
|
| 96 |
) and f.endswith(".txt"):
|
| 97 |
path = os.path.join(TMP_DIR, f)
|
| 98 |
try:
|
|
@@ -138,6 +140,7 @@ def remove_session_hash_code_data(session_hash_code: str):
|
|
| 138 |
# --- Define all possible session_hash_code file patterns ---
|
| 139 |
files_to_remove = [
|
| 140 |
get_active_task_flag_file(session_hash_code),
|
|
|
|
| 141 |
]
|
| 142 |
|
| 143 |
# --- Remove all temporary files ---
|
|
@@ -164,11 +167,6 @@ def remove_session_hash_code_data(session_hash_code: str):
|
|
| 164 |
except Exception as e:
|
| 165 |
logging.error(f"[{session_hash_code}] Error during reset_session: {e}")
|
| 166 |
|
| 167 |
-
def generate_session_id() -> str:
|
| 168 |
-
"""Generates a unique session_hash_code ID."""
|
| 169 |
-
sid = str(uuid.uuid4())
|
| 170 |
-
logging.debug(f"[{sid}] New session_hash_code created.")
|
| 171 |
-
return sid
|
| 172 |
|
| 173 |
|
| 174 |
def register_session_hash_code(session_hash_code: str, filepath: str):
|
|
@@ -239,6 +237,27 @@ def get_active_session_hash_code():
|
|
| 239 |
def get_active_task_flag_file(session_hash_code: str):
|
| 240 |
return os.path.join(TMP_DIR, f"{ACTIVE_TASK_FLAG}{session_hash_code}.txt")
|
| 241 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 242 |
|
| 243 |
def get_folder_chunks(session_hash_code: str):
|
| 244 |
return os.path.join(TMP_DIR, f"{NAME_FOLDER_CHUNKS}{session_hash_code}")
|
|
|
|
| 9 |
|
| 10 |
TMP_DIR = os.getenv("TMP_DIR", "/tmp/canary_aed_streaming")
|
| 11 |
ACTIVE_SESSIONS_HASH_FILE = os.path.join(TMP_DIR, "active_session_hash_code.json")
|
| 12 |
+
ACTIVE_STREAM_FLAG="stream_active_"
|
| 13 |
ACTIVE_TASK_FLAG="task_active_"
|
| 14 |
NAME_FOLDER_CHUNKS="chunks_"
|
| 15 |
|
|
|
|
| 59 |
# ---------------------------
|
| 60 |
def on_unload(request: gr.Request):
|
| 61 |
"""Called when the visitor closes or refreshes the app."""
|
| 62 |
+
session_hash_code = request.session_hash
|
| 63 |
sessions = _read_session_hash_code()
|
| 64 |
|
| 65 |
+
if session_hash_code in sessions:
|
| 66 |
+
sessions.pop(session_hash_code)
|
| 67 |
_write_session_hash_code(sessions)
|
| 68 |
+
remove_session_hash_code_data(session_hash_code)
|
| 69 |
+
unregister_session_hash_code_hash(session_hash_code)
|
| 70 |
+
logging.info(f"[{session_hash_code}] session_hash_code removed (on_unload).")
|
| 71 |
else:
|
| 72 |
+
logging.info(f"[{session_hash_code}] No active session_hash_code found to remove.")
|
| 73 |
|
| 74 |
def ensure_tmp_dir():
|
| 75 |
"""Ensures the base temporary directory exists."""
|
|
|
|
| 93 |
# --- Clean all flag files (stream + transcribe) ---
|
| 94 |
for f in os.listdir(TMP_DIR):
|
| 95 |
if (
|
| 96 |
+
f.startswith(f"{ACTIVE_TASK_FLAG}")
|
| 97 |
+
or f.startswith(f"{ACTIVE_STREAM_FLAG}")
|
| 98 |
) and f.endswith(".txt"):
|
| 99 |
path = os.path.join(TMP_DIR, f)
|
| 100 |
try:
|
|
|
|
| 140 |
# --- Define all possible session_hash_code file patterns ---
|
| 141 |
files_to_remove = [
|
| 142 |
get_active_task_flag_file(session_hash_code),
|
| 143 |
+
get_active_stream_flag_file(session_hash_code),
|
| 144 |
]
|
| 145 |
|
| 146 |
# --- Remove all temporary files ---
|
|
|
|
| 167 |
except Exception as e:
|
| 168 |
logging.error(f"[{session_hash_code}] Error during reset_session: {e}")
|
| 169 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 170 |
|
| 171 |
|
| 172 |
def register_session_hash_code(session_hash_code: str, filepath: str):
|
|
|
|
| 237 |
def get_active_task_flag_file(session_hash_code: str):
|
| 238 |
return os.path.join(TMP_DIR, f"{ACTIVE_TASK_FLAG}{session_hash_code}.txt")
|
| 239 |
|
| 240 |
+
def get_active_stream_flag_file(session_hash_code: str):
|
| 241 |
+
return os.path.join(TMP_DIR, f"{ACTIVE_STREAM_FLAG}{session_hash_code}.txt")
|
| 242 |
+
|
| 243 |
+
|
| 244 |
+
def remove_active_stream_flag_file(session_hash_code: str):
|
| 245 |
+
fname = os.path.join(TMP_DIR, f"{ACTIVE_STREAM_FLAG}{session_hash_code}.txt")
|
| 246 |
+
if os.path.exists(fname):
|
| 247 |
+
try:
|
| 248 |
+
os.remove(fname)
|
| 249 |
+
logging.debug(f"[{session_hash_code}] Removed file: {fname}")
|
| 250 |
+
except Exception as e:
|
| 251 |
+
logging.warning(f"[{session_hash_code}] Failed to remove file {fname}: {e}")
|
| 252 |
+
|
| 253 |
+
def remove_active_task_flag_file(session_hash_code: str):
|
| 254 |
+
fname = os.path.join(TMP_DIR, f"{ACTIVE_TASK_FLAG}{session_hash_code}.txt")
|
| 255 |
+
if os.path.exists(fname):
|
| 256 |
+
try:
|
| 257 |
+
os.remove(fname)
|
| 258 |
+
logging.debug(f"[{session_hash_code}] Removed file: {fname}")
|
| 259 |
+
except Exception as e:
|
| 260 |
+
logging.warning(f"[{session_hash_code}] Failed to remove file {fname}: {e}")
|
| 261 |
|
| 262 |
def get_folder_chunks(session_hash_code: str):
|
| 263 |
return os.path.join(TMP_DIR, f"{NAME_FOLDER_CHUNKS}{session_hash_code}")
|
app/streaming_audio_processor.py
CHANGED
|
@@ -84,18 +84,19 @@ class StreamingAudioProcessor:
|
|
| 84 |
Flushes the remaining buffer to the transcriber, resets the state,
|
| 85 |
and returns the last transcribed text.
|
| 86 |
"""
|
| 87 |
-
new_text = ""
|
| 88 |
if len(self.internal_buffer) > 0:
|
| 89 |
# Buffer is already a numpy array
|
| 90 |
final_segment_chunk = self.internal_buffer
|
| 91 |
logging.info(f"Flushing segment remainder of {len(final_segment_chunk)} samples.")
|
| 92 |
-
seg, new_text
|
|
|
|
| 93 |
else:
|
| 94 |
# Buffer is empty, but send a silent "flush"
|
| 95 |
# to force the transcriber to finalize its internal state.
|
| 96 |
logging.info("Buffer empty, sending silent flush to finalize segment.")
|
| 97 |
flush_chunk = np.zeros(self.logical_chunk_size, dtype='int16')
|
| 98 |
-
seg, new_text
|
|
|
|
| 99 |
|
| 100 |
# Full state reset
|
| 101 |
logging.debug("Resetting speech engine state...")
|
|
@@ -106,7 +107,7 @@ class StreamingAudioProcessor:
|
|
| 106 |
self.is_first_logical_chunk = True
|
| 107 |
self.silent_chunks_count = 0
|
| 108 |
|
| 109 |
-
|
| 110 |
|
| 111 |
def process_chunk(self, chunk: np.ndarray):
|
| 112 |
"""
|
|
@@ -142,10 +143,12 @@ class StreamingAudioProcessor:
|
|
| 142 |
|
| 143 |
if asr_chunk_np is not None:
|
| 144 |
logging.debug(f"Sending logical chunk (size: {len(asr_chunk_np)}) to speech engine...")
|
| 145 |
-
seg, new_text
|
| 146 |
-
if new_text:
|
| 147 |
logging.info(f"Received new text segment: '{new_text}'")
|
| 148 |
new_text_segments.append(new_text)
|
|
|
|
|
|
|
|
|
|
| 149 |
self.is_first_logical_chunk = False
|
| 150 |
|
| 151 |
# --- 3. VAD Reset Logic ---
|
|
@@ -153,12 +156,14 @@ class StreamingAudioProcessor:
|
|
| 153 |
logging.info(f"\n[VAD RESET: SILENCE detected ({self.silent_chunks_count} empty chunks) at {(self.chunks_count * (self.read_size/self.VAD_SAMPLE_RATE)):.2f}s]")
|
| 154 |
|
| 155 |
# Flush the buffer, reset state, and get final text
|
| 156 |
-
reset_text
|
| 157 |
-
if reset_text:
|
| 158 |
logging.info(f"Received final reset text: '{reset_text}'")
|
| 159 |
new_text_segments.append(reset_text)
|
| 160 |
-
|
| 161 |
-
|
|
|
|
|
|
|
|
|
|
| 162 |
|
| 163 |
def finalize_stream(self):
|
| 164 |
"""
|
|
|
|
| 84 |
Flushes the remaining buffer to the transcriber, resets the state,
|
| 85 |
and returns the last transcribed text.
|
| 86 |
"""
|
|
|
|
| 87 |
if len(self.internal_buffer) > 0:
|
| 88 |
# Buffer is already a numpy array
|
| 89 |
final_segment_chunk = self.internal_buffer
|
| 90 |
logging.info(f"Flushing segment remainder of {len(final_segment_chunk)} samples.")
|
| 91 |
+
for seg, new_text in self.speech_engine.transcribe_chunk(final_segment_chunk, is_last_chunk=True) :
|
| 92 |
+
yield new_text
|
| 93 |
else:
|
| 94 |
# Buffer is empty, but send a silent "flush"
|
| 95 |
# to force the transcriber to finalize its internal state.
|
| 96 |
logging.info("Buffer empty, sending silent flush to finalize segment.")
|
| 97 |
flush_chunk = np.zeros(self.logical_chunk_size, dtype='int16')
|
| 98 |
+
for seg, new_text in self.speech_engine.transcribe_chunk(flush_chunk, is_last_chunk=True) :
|
| 99 |
+
yield new_text
|
| 100 |
|
| 101 |
# Full state reset
|
| 102 |
logging.debug("Resetting speech engine state...")
|
|
|
|
| 107 |
self.is_first_logical_chunk = True
|
| 108 |
self.silent_chunks_count = 0
|
| 109 |
|
| 110 |
+
yield ""
|
| 111 |
|
| 112 |
def process_chunk(self, chunk: np.ndarray):
|
| 113 |
"""
|
|
|
|
| 143 |
|
| 144 |
if asr_chunk_np is not None:
|
| 145 |
logging.debug(f"Sending logical chunk (size: {len(asr_chunk_np)}) to speech engine...")
|
| 146 |
+
for seg, new_text in self.speech_engine.transcribe_chunk(asr_chunk_np, is_last_chunk=False) :
|
|
|
|
| 147 |
logging.info(f"Received new text segment: '{new_text}'")
|
| 148 |
new_text_segments.append(new_text)
|
| 149 |
+
yield new_text
|
| 150 |
+
else :
|
| 151 |
+
yield ""
|
| 152 |
self.is_first_logical_chunk = False
|
| 153 |
|
| 154 |
# --- 3. VAD Reset Logic ---
|
|
|
|
| 156 |
logging.info(f"\n[VAD RESET: SILENCE detected ({self.silent_chunks_count} empty chunks) at {(self.chunks_count * (self.read_size/self.VAD_SAMPLE_RATE)):.2f}s]")
|
| 157 |
|
| 158 |
# Flush the buffer, reset state, and get final text
|
| 159 |
+
for reset_text in self._flush_and_reset() :
|
|
|
|
| 160 |
logging.info(f"Received final reset text: '{reset_text}'")
|
| 161 |
new_text_segments.append(reset_text)
|
| 162 |
+
yield reset_text
|
| 163 |
+
else :
|
| 164 |
+
yield ""
|
| 165 |
+
|
| 166 |
+
yield ""
|
| 167 |
|
| 168 |
def finalize_stream(self):
|
| 169 |
"""
|
app/utils.py
CHANGED
|
@@ -16,6 +16,9 @@ import torch
|
|
| 16 |
from app.streaming_audio_processor import StreamingAudioProcessor
|
| 17 |
from app.session_utils import (
|
| 18 |
get_active_task_flag_file,
|
|
|
|
|
|
|
|
|
|
| 19 |
get_folder_chunks
|
| 20 |
)
|
| 21 |
from app.ui_utils import (
|
|
@@ -65,73 +68,73 @@ def generate_coturn_config():
|
|
| 65 |
|
| 66 |
|
| 67 |
|
| 68 |
-
def read_and_stream_audio(filepath_to_stream: str,
|
| 69 |
"""
|
| 70 |
Read an audio file and stream it chunk by chunk (1s per chunk).
|
| 71 |
Handles errors safely and reports structured messages to the client.
|
| 72 |
"""
|
| 73 |
-
if not
|
| 74 |
-
yield from handle_stream_error("unknown", "No
|
| 75 |
return
|
| 76 |
|
| 77 |
if not filepath_to_stream or not os.path.exists(filepath_to_stream):
|
| 78 |
-
yield from handle_stream_error(
|
| 79 |
return
|
| 80 |
-
|
| 81 |
try:
|
| 82 |
segment = AudioSegment.from_file(filepath_to_stream)
|
| 83 |
chunk_duration_ms = int((read_size/sample_rate)*1000)
|
| 84 |
total_chunks = len(segment) // chunk_duration_ms + 1
|
| 85 |
-
|
|
|
|
| 86 |
|
| 87 |
for i, chunk in enumerate(segment[::chunk_duration_ms]):
|
| 88 |
-
|
| 89 |
|
| 90 |
frame_rate = chunk.frame_rate
|
| 91 |
samples = np.array(chunk.get_array_of_samples()).reshape(1, -1)
|
| 92 |
progress = round(((i + 1) / total_chunks) * 100, 2)
|
| 93 |
-
if _is_stop_requested(
|
| 94 |
-
logging.info(f"[{
|
| 95 |
yield ((frame_rate, samples), AdditionalOutputs({"stoped": True, "value": "STREAM_STOPED"} ) )
|
| 96 |
break
|
| 97 |
|
| 98 |
yield ((frame_rate, samples), AdditionalOutputs({"progressed": True, "value": progress} ))
|
| 99 |
-
|
| 100 |
|
| 101 |
time.sleep(chunk_duration_ms/1000)
|
| 102 |
# Save only if transcription is active
|
| 103 |
-
if os.path.exists(
|
| 104 |
-
chunk_dir = get_folder_chunks(
|
| 105 |
if not os.path.exists(chunk_dir) :
|
| 106 |
os.makedirs(chunk_dir, exist_ok=True)
|
| 107 |
npz_path = os.path.join(chunk_dir, f"chunk_{i:05d}.npz")
|
| 108 |
chunk_array = np.array(chunk.get_array_of_samples(), dtype=np.int16)
|
| 109 |
-
|
| 110 |
-
|
|
|
|
| 111 |
|
| 112 |
# raise_function() # Optional injected test exception
|
| 113 |
|
| 114 |
-
logging.info(f"[{
|
| 115 |
|
| 116 |
except asyncio.CancelledError:
|
| 117 |
-
yield from handle_stream_error(
|
| 118 |
except FileNotFoundError as e:
|
| 119 |
-
yield from handle_stream_error(
|
| 120 |
except Exception as e:
|
| 121 |
-
yield from handle_stream_error(
|
| 122 |
-
|
| 123 |
finally:
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
logging.info(f"[{session_id}] Stop flag reset.")
|
| 127 |
|
| 128 |
|
| 129 |
|
| 130 |
-
|
| 131 |
-
asr_model = None
|
| 132 |
|
| 133 |
@spaces.GPU
|
| 134 |
-
def task_fake(
|
| 135 |
task_type, lang_source, lang_target,
|
| 136 |
chunk_secs, left_context_secs, right_context_secs,
|
| 137 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
@@ -158,21 +161,21 @@ def task_fake(session_id: str,
|
|
| 158 |
# streamer = StreamingAudioProcessor(speech_engine=canary_speech_engine,vad_engine=silero_vad_engine,cfg=streaming_audio_processor_config)
|
| 159 |
##-----------
|
| 160 |
yield ("initialized the CanarySpeechEngine and Silero_Vad_Engine", "info", None)
|
| 161 |
-
yield (f"Task started for session {
|
| 162 |
|
| 163 |
-
active_flag = get_active_task_flag_file(
|
| 164 |
with open(active_flag, "w") as f:
|
| 165 |
f.write("1")
|
| 166 |
-
chunk_dir = get_folder_chunks(
|
| 167 |
-
logging.info(f"[{
|
| 168 |
|
| 169 |
try:
|
| 170 |
-
logging.info(f"[{
|
| 171 |
-
yield (f"Task started for session {
|
| 172 |
|
| 173 |
while os.path.exists(active_flag):
|
| 174 |
if not os.path.exists(chunk_dir):
|
| 175 |
-
logging.warning(f"[{
|
| 176 |
yield ("No audio chunks yet... waiting for stream.", "warning", None)
|
| 177 |
time.sleep(0.1)
|
| 178 |
continue
|
|
@@ -193,15 +196,15 @@ def task_fake(session_id: str,
|
|
| 193 |
# for text in new_texts:
|
| 194 |
# print(text, end='', flush=True)
|
| 195 |
# yield (text, "success", text)
|
| 196 |
-
# logging.debug(f"[{
|
| 197 |
##-----------
|
| 198 |
### TODO
|
| 199 |
text = f"Transcribed {fname}: {len(samples)} samples @ {rate}Hz\n"
|
| 200 |
yield (text, "success", fname)
|
| 201 |
os.remove(fpath)
|
| 202 |
-
logging.debug(f"[{
|
| 203 |
except Exception as e:
|
| 204 |
-
logging.warning(f"[{
|
| 205 |
yield (f"Error processing {fname}: {e}", "warning", fname)
|
| 206 |
continue
|
| 207 |
time.sleep(0.1)
|
|
@@ -212,31 +215,31 @@ def task_fake(session_id: str,
|
|
| 212 |
# yield (text, "success", final_text)
|
| 213 |
##-----------
|
| 214 |
yield ("DONE", "done", None)
|
| 215 |
-
logging.info(f"[{
|
| 216 |
|
| 217 |
except Exception as e:
|
| 218 |
-
logging.error(f"[{
|
| 219 |
yield (f"Unexpected error: {e}", "error", None)
|
| 220 |
|
| 221 |
finally:
|
| 222 |
if os.path.exists(active_flag):
|
| 223 |
os.remove(active_flag)
|
| 224 |
-
logging.info(f"[{
|
| 225 |
|
| 226 |
try:
|
| 227 |
if os.path.exists(chunk_dir) and not os.listdir(chunk_dir):
|
| 228 |
os.rmdir(chunk_dir)
|
| 229 |
-
logging.debug(f"[{
|
| 230 |
except Exception as e:
|
| 231 |
-
logging.error(f"[{
|
| 232 |
yield (f"Cleanup error: {e}", "error", None)
|
| 233 |
|
| 234 |
-
logging.info(f"[{
|
| 235 |
yield ("Task finished and cleaned up.", "done", None)
|
| 236 |
|
| 237 |
|
| 238 |
-
|
| 239 |
-
def task(
|
| 240 |
task_type, lang_source, lang_target,
|
| 241 |
chunk_secs, left_context_secs, right_context_secs,
|
| 242 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
@@ -260,21 +263,21 @@ def task(session_id: str,
|
|
| 260 |
)
|
| 261 |
streamer = StreamingAudioProcessor(speech_engine=canary_speech_engine,vad_engine=silero_vad_engine,cfg=streaming_audio_processor_config)
|
| 262 |
yield ("initialized the CanarySpeechEngine and Silero_Vad_Engine", "info", None)
|
| 263 |
-
yield (f"Task started for session {
|
| 264 |
|
| 265 |
-
active_flag = get_active_task_flag_file(
|
| 266 |
with open(active_flag, "w") as f:
|
| 267 |
f.write("1")
|
| 268 |
-
chunk_dir = get_folder_chunks(
|
| 269 |
-
logging.info(f"[{
|
| 270 |
|
| 271 |
try:
|
| 272 |
-
logging.info(f"[{
|
| 273 |
-
yield (f"Task started for session {
|
| 274 |
|
| 275 |
while os.path.exists(active_flag):
|
| 276 |
if not os.path.exists(chunk_dir):
|
| 277 |
-
logging.warning(f"[{
|
| 278 |
yield ("No audio chunks yet... waiting for stream.", "warning", None)
|
| 279 |
time.sleep(0.1)
|
| 280 |
continue
|
|
@@ -290,54 +293,52 @@ def task(session_id: str,
|
|
| 290 |
npz = np.load(fpath)
|
| 291 |
samples = npz["data"]
|
| 292 |
rate = int(npz["rate"])
|
| 293 |
-
|
| 294 |
-
for text in new_texts:
|
| 295 |
-
print(text, end='', flush=True)
|
| 296 |
yield (text, "success", text)
|
| 297 |
-
logging.debug(f"[{
|
| 298 |
### TODO
|
| 299 |
# text = f"Transcribed {fname}: {len(samples)} samples @ {rate}Hz\n"
|
| 300 |
# yield (text, "success", fname)
|
| 301 |
os.remove(fpath)
|
| 302 |
-
logging.debug(f"[{
|
| 303 |
except Exception as e:
|
| 304 |
-
logging.warning(f"[{
|
| 305 |
yield (f"Error processing {fname}: {e}", "warning", fname)
|
| 306 |
continue
|
| 307 |
time.sleep(0.1)
|
| 308 |
|
| 309 |
# TODO
|
| 310 |
-
final_text = streamer.finalize_stream()
|
| 311 |
-
yield (text, "success", final_text)
|
| 312 |
# if final_text:
|
| 313 |
# print(final_text, end='', flush=True)
|
| 314 |
# yield f"\n{final_text}"
|
| 315 |
##
|
| 316 |
yield ("DONE", "done", None)
|
| 317 |
-
logging.info(f"[{
|
| 318 |
|
| 319 |
except Exception as e:
|
| 320 |
-
logging.error(f"[{
|
| 321 |
yield (f"Unexpected error: {e}", "error", None)
|
| 322 |
|
| 323 |
finally:
|
| 324 |
if os.path.exists(active_flag):
|
| 325 |
os.remove(active_flag)
|
| 326 |
-
logging.info(f"[{
|
| 327 |
|
| 328 |
try:
|
| 329 |
if os.path.exists(chunk_dir) and not os.listdir(chunk_dir):
|
| 330 |
os.rmdir(chunk_dir)
|
| 331 |
-
logging.debug(f"[{
|
| 332 |
except Exception as e:
|
| 333 |
-
logging.error(f"[{
|
| 334 |
yield (f"Cleanup error: {e}", "error", None)
|
| 335 |
|
| 336 |
-
logging.info(f"[{
|
| 337 |
yield ("Task finished and cleaned up.", "done", None)
|
| 338 |
|
| 339 |
|
| 340 |
-
def handle_stream_error(
|
| 341 |
"""
|
| 342 |
Handle streaming errors:
|
| 343 |
- Log the error
|
|
@@ -349,20 +350,14 @@ def handle_stream_error(session_id: str, error: Exception | str, stop_streaming_
|
|
| 349 |
else:
|
| 350 |
msg = str(error)
|
| 351 |
|
| 352 |
-
logging.error(f"[{
|
| 353 |
|
| 354 |
-
|
| 355 |
-
stop_streaming_flags["stop"] = False
|
| 356 |
|
| 357 |
yield ((16000,np.zeros(16000, dtype=np.float32).reshape(1, -1)), AdditionalOutputs({"errored": True, "value": msg}))
|
| 358 |
|
| 359 |
|
| 360 |
|
| 361 |
-
def _is_stop_requested(stop_streaming_flags: dict) -> bool:
|
| 362 |
-
"""Check if the stop signal was requested."""
|
| 363 |
-
if not isinstance(stop_streaming_flags, dict):
|
| 364 |
-
return False
|
| 365 |
-
return bool(stop_streaming_flags.get("stop", False))
|
| 366 |
|
| 367 |
# --- Decorator compatibility layer ---
|
| 368 |
if os.environ.get("SPACE_ID", "").startswith("zero-gpu"):
|
|
@@ -378,14 +373,23 @@ else:
|
|
| 378 |
|
| 379 |
|
| 380 |
|
| 381 |
-
def stop_streaming(
|
| 382 |
"""Trigger the stop flag for active streaming."""
|
| 383 |
-
logging.info(f"[{
|
| 384 |
-
|
| 385 |
-
|
| 386 |
-
|
| 387 |
-
|
| 388 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 389 |
|
| 390 |
|
| 391 |
def raise_function():
|
|
|
|
| 16 |
from app.streaming_audio_processor import StreamingAudioProcessor
|
| 17 |
from app.session_utils import (
|
| 18 |
get_active_task_flag_file,
|
| 19 |
+
get_active_stream_flag_file,
|
| 20 |
+
remove_active_stream_flag_file,
|
| 21 |
+
remove_active_task_flag_file,
|
| 22 |
get_folder_chunks
|
| 23 |
)
|
| 24 |
from app.ui_utils import (
|
|
|
|
| 68 |
|
| 69 |
|
| 70 |
|
| 71 |
+
def read_and_stream_audio(filepath_to_stream: str, session_hash_code: str,read_size:int =8000, sample_rate:int =16000):
|
| 72 |
"""
|
| 73 |
Read an audio file and stream it chunk by chunk (1s per chunk).
|
| 74 |
Handles errors safely and reports structured messages to the client.
|
| 75 |
"""
|
| 76 |
+
if not session_hash_code:
|
| 77 |
+
yield from handle_stream_error("unknown", "No session_hash_code provided.")
|
| 78 |
return
|
| 79 |
|
| 80 |
if not filepath_to_stream or not os.path.exists(filepath_to_stream):
|
| 81 |
+
yield from handle_stream_error(session_hash_code, f"Audio file not found: {filepath_to_stream}")
|
| 82 |
return
|
| 83 |
+
task_active_flag = get_active_task_flag_file(session_hash_code)
|
| 84 |
try:
|
| 85 |
segment = AudioSegment.from_file(filepath_to_stream)
|
| 86 |
chunk_duration_ms = int((read_size/sample_rate)*1000)
|
| 87 |
total_chunks = len(segment) // chunk_duration_ms + 1
|
| 88 |
+
start_streaming(session_hash_code)
|
| 89 |
+
logging.info(f"[{session_hash_code}] Starting audio streaming {filepath_to_stream} ({total_chunks} chunks).")
|
| 90 |
|
| 91 |
for i, chunk in enumerate(segment[::chunk_duration_ms]):
|
| 92 |
+
|
| 93 |
|
| 94 |
frame_rate = chunk.frame_rate
|
| 95 |
samples = np.array(chunk.get_array_of_samples()).reshape(1, -1)
|
| 96 |
progress = round(((i + 1) / total_chunks) * 100, 2)
|
| 97 |
+
if _is_stop_requested(session_hash_code):
|
| 98 |
+
logging.info(f"[{session_hash_code}] Stop signal received. Terminating stream.")
|
| 99 |
yield ((frame_rate, samples), AdditionalOutputs({"stoped": True, "value": "STREAM_STOPED"} ) )
|
| 100 |
break
|
| 101 |
|
| 102 |
yield ((frame_rate, samples), AdditionalOutputs({"progressed": True, "value": progress} ))
|
| 103 |
+
logging.debug(f"[{session_hash_code}] Sent chunk {i+1}/{total_chunks} ({progress}%).")
|
| 104 |
|
| 105 |
time.sleep(chunk_duration_ms/1000)
|
| 106 |
# Save only if transcription is active
|
| 107 |
+
if os.path.exists(task_active_flag) :
|
| 108 |
+
chunk_dir = get_folder_chunks(session_hash_code)
|
| 109 |
if not os.path.exists(chunk_dir) :
|
| 110 |
os.makedirs(chunk_dir, exist_ok=True)
|
| 111 |
npz_path = os.path.join(chunk_dir, f"chunk_{i:05d}.npz")
|
| 112 |
chunk_array = np.array(chunk.get_array_of_samples(), dtype=np.int16)
|
| 113 |
+
if os.path.exists(task_active_flag):
|
| 114 |
+
np.savez_compressed(npz_path, data=chunk_array, rate=frame_rate)
|
| 115 |
+
logging.debug(f"[{session_hash_code}] Saved chunk {i}/{total_chunks} (transcribe active) ({progress}%) ({npz_path}).")
|
| 116 |
|
| 117 |
# raise_function() # Optional injected test exception
|
| 118 |
|
| 119 |
+
logging.info(f"[{session_hash_code}] Audio streaming completed successfully.")
|
| 120 |
|
| 121 |
except asyncio.CancelledError:
|
| 122 |
+
yield from handle_stream_error(session_hash_code, "Streaming cancelled by user.")
|
| 123 |
except FileNotFoundError as e:
|
| 124 |
+
yield from handle_stream_error(session_hash_code, e)
|
| 125 |
except Exception as e:
|
| 126 |
+
yield from handle_stream_error(session_hash_code, e)
|
|
|
|
| 127 |
finally:
|
| 128 |
+
remove_active_stream_flag_file(session_hash_code)
|
| 129 |
+
logging.info(f"[{session_hash_code}] Stop flag reset.")
|
|
|
|
| 130 |
|
| 131 |
|
| 132 |
|
| 133 |
+
asr_model = nemo_asr.models.ASRModel.from_pretrained("nvidia/canary-1b-v2")
|
| 134 |
+
# asr_model = None
|
| 135 |
|
| 136 |
@spaces.GPU
|
| 137 |
+
def task_fake(session_hash_code: str,
|
| 138 |
task_type, lang_source, lang_target,
|
| 139 |
chunk_secs, left_context_secs, right_context_secs,
|
| 140 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
|
|
| 161 |
# streamer = StreamingAudioProcessor(speech_engine=canary_speech_engine,vad_engine=silero_vad_engine,cfg=streaming_audio_processor_config)
|
| 162 |
##-----------
|
| 163 |
yield ("initialized the CanarySpeechEngine and Silero_Vad_Engine", "info", None)
|
| 164 |
+
yield (f"Task started for session {session_hash_code}", "info", None)
|
| 165 |
|
| 166 |
+
active_flag = get_active_task_flag_file(session_hash_code)
|
| 167 |
with open(active_flag, "w") as f:
|
| 168 |
f.write("1")
|
| 169 |
+
chunk_dir = get_folder_chunks(session_hash_code)
|
| 170 |
+
logging.info(f"[{session_hash_code}] task started. {chunk_dir}")
|
| 171 |
|
| 172 |
try:
|
| 173 |
+
logging.info(f"[{session_hash_code}] task loop started.")
|
| 174 |
+
yield (f"Task started for session {session_hash_code}", "info", None)
|
| 175 |
|
| 176 |
while os.path.exists(active_flag):
|
| 177 |
if not os.path.exists(chunk_dir):
|
| 178 |
+
logging.warning(f"[{session_hash_code}] No chunk directory found for task.")
|
| 179 |
yield ("No audio chunks yet... waiting for stream.", "warning", None)
|
| 180 |
time.sleep(0.1)
|
| 181 |
continue
|
|
|
|
| 196 |
# for text in new_texts:
|
| 197 |
# print(text, end='', flush=True)
|
| 198 |
# yield (text, "success", text)
|
| 199 |
+
# logging.debug(f"[{session_hash_code}] {new_texts}")
|
| 200 |
##-----------
|
| 201 |
### TODO
|
| 202 |
text = f"Transcribed {fname}: {len(samples)} samples @ {rate}Hz\n"
|
| 203 |
yield (text, "success", fname)
|
| 204 |
os.remove(fpath)
|
| 205 |
+
logging.debug(f"[{session_hash_code}] Deleted processed chunk: {fname}")
|
| 206 |
except Exception as e:
|
| 207 |
+
logging.warning(f"[{session_hash_code}] Error processing {fname}: {e}")
|
| 208 |
yield (f"Error processing {fname}: {e}", "warning", fname)
|
| 209 |
continue
|
| 210 |
time.sleep(0.1)
|
|
|
|
| 215 |
# yield (text, "success", final_text)
|
| 216 |
##-----------
|
| 217 |
yield ("DONE", "done", None)
|
| 218 |
+
logging.info(f"[{session_hash_code}] task loop ended (flag removed).")
|
| 219 |
|
| 220 |
except Exception as e:
|
| 221 |
+
logging.error(f"[{session_hash_code}] task error: {e}", exc_info=True)
|
| 222 |
yield (f"Unexpected error: {e}", "error", None)
|
| 223 |
|
| 224 |
finally:
|
| 225 |
if os.path.exists(active_flag):
|
| 226 |
os.remove(active_flag)
|
| 227 |
+
logging.info(f"[{session_hash_code}] task stopped.")
|
| 228 |
|
| 229 |
try:
|
| 230 |
if os.path.exists(chunk_dir) and not os.listdir(chunk_dir):
|
| 231 |
os.rmdir(chunk_dir)
|
| 232 |
+
logging.debug(f"[{session_hash_code}] Cleaned up empty chunk dir.")
|
| 233 |
except Exception as e:
|
| 234 |
+
logging.error(f"[{session_hash_code}] Cleanup error: {e}")
|
| 235 |
yield (f"Cleanup error: {e}", "error", None)
|
| 236 |
|
| 237 |
+
logging.info(f"[{session_hash_code}] Exiting task loop.")
|
| 238 |
yield ("Task finished and cleaned up.", "done", None)
|
| 239 |
|
| 240 |
|
| 241 |
+
@spaces.GPU
|
| 242 |
+
def task(session_hash_code: str,
|
| 243 |
task_type, lang_source, lang_target,
|
| 244 |
chunk_secs, left_context_secs, right_context_secs,
|
| 245 |
streaming_policy, alignatt_thr, waitk_lagging,
|
|
|
|
| 263 |
)
|
| 264 |
streamer = StreamingAudioProcessor(speech_engine=canary_speech_engine,vad_engine=silero_vad_engine,cfg=streaming_audio_processor_config)
|
| 265 |
yield ("initialized the CanarySpeechEngine and Silero_Vad_Engine", "info", None)
|
| 266 |
+
yield (f"Task started for session {session_hash_code}", "info", None)
|
| 267 |
|
| 268 |
+
active_flag = get_active_task_flag_file(session_hash_code)
|
| 269 |
with open(active_flag, "w") as f:
|
| 270 |
f.write("1")
|
| 271 |
+
chunk_dir = get_folder_chunks(session_hash_code)
|
| 272 |
+
logging.info(f"[{session_hash_code}] task started. {chunk_dir}")
|
| 273 |
|
| 274 |
try:
|
| 275 |
+
logging.info(f"[{session_hash_code}] task loop started.")
|
| 276 |
+
yield (f"Task started for session {session_hash_code}", "info", None)
|
| 277 |
|
| 278 |
while os.path.exists(active_flag):
|
| 279 |
if not os.path.exists(chunk_dir):
|
| 280 |
+
logging.warning(f"[{session_hash_code}] No chunk directory found for task.")
|
| 281 |
yield ("No audio chunks yet... waiting for stream.", "warning", None)
|
| 282 |
time.sleep(0.1)
|
| 283 |
continue
|
|
|
|
| 293 |
npz = np.load(fpath)
|
| 294 |
samples = npz["data"]
|
| 295 |
rate = int(npz["rate"])
|
| 296 |
+
for text in streamer.process_chunk(samples) :
|
|
|
|
|
|
|
| 297 |
yield (text, "success", text)
|
| 298 |
+
logging.debug(f"[{session_hash_code}] {text}")
|
| 299 |
### TODO
|
| 300 |
# text = f"Transcribed {fname}: {len(samples)} samples @ {rate}Hz\n"
|
| 301 |
# yield (text, "success", fname)
|
| 302 |
os.remove(fpath)
|
| 303 |
+
logging.debug(f"[{session_hash_code}] Deleted processed chunk: {fname}")
|
| 304 |
except Exception as e:
|
| 305 |
+
logging.warning(f"[{session_hash_code}] Error processing {fname}: {e}")
|
| 306 |
yield (f"Error processing {fname}: {e}", "warning", fname)
|
| 307 |
continue
|
| 308 |
time.sleep(0.1)
|
| 309 |
|
| 310 |
# TODO
|
| 311 |
+
# final_text = streamer.finalize_stream()
|
| 312 |
+
# yield (text, "success", final_text)
|
| 313 |
# if final_text:
|
| 314 |
# print(final_text, end='', flush=True)
|
| 315 |
# yield f"\n{final_text}"
|
| 316 |
##
|
| 317 |
yield ("DONE", "done", None)
|
| 318 |
+
logging.info(f"[{session_hash_code}] task loop ended (flag removed).")
|
| 319 |
|
| 320 |
except Exception as e:
|
| 321 |
+
logging.error(f"[{session_hash_code}] task error: {e}", exc_info=True)
|
| 322 |
yield (f"Unexpected error: {e}", "error", None)
|
| 323 |
|
| 324 |
finally:
|
| 325 |
if os.path.exists(active_flag):
|
| 326 |
os.remove(active_flag)
|
| 327 |
+
logging.info(f"[{session_hash_code}] task stopped.")
|
| 328 |
|
| 329 |
try:
|
| 330 |
if os.path.exists(chunk_dir) and not os.listdir(chunk_dir):
|
| 331 |
os.rmdir(chunk_dir)
|
| 332 |
+
logging.debug(f"[{session_hash_code}] Cleaned up empty chunk dir.")
|
| 333 |
except Exception as e:
|
| 334 |
+
logging.error(f"[{session_hash_code}] Cleanup error: {e}")
|
| 335 |
yield (f"Cleanup error: {e}", "error", None)
|
| 336 |
|
| 337 |
+
logging.info(f"[{session_hash_code}] Exiting task loop.")
|
| 338 |
yield ("Task finished and cleaned up.", "done", None)
|
| 339 |
|
| 340 |
|
| 341 |
+
def handle_stream_error(session_hash_code: str, error: Exception):
|
| 342 |
"""
|
| 343 |
Handle streaming errors:
|
| 344 |
- Log the error
|
|
|
|
| 350 |
else:
|
| 351 |
msg = str(error)
|
| 352 |
|
| 353 |
+
logging.error(f"[{session_hash_code}] Streaming error: {msg}", exc_info=isinstance(error, Exception))
|
| 354 |
|
| 355 |
+
remove_active_stream_flag_file(session_hash_code)
|
|
|
|
| 356 |
|
| 357 |
yield ((16000,np.zeros(16000, dtype=np.float32).reshape(1, -1)), AdditionalOutputs({"errored": True, "value": msg}))
|
| 358 |
|
| 359 |
|
| 360 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 361 |
|
| 362 |
# --- Decorator compatibility layer ---
|
| 363 |
if os.environ.get("SPACE_ID", "").startswith("zero-gpu"):
|
|
|
|
| 373 |
|
| 374 |
|
| 375 |
|
| 376 |
+
def stop_streaming(session_hash_code: str):
|
| 377 |
"""Trigger the stop flag for active streaming."""
|
| 378 |
+
logging.info(f"[{session_hash_code}] Stop button clicked — sending stop signal.")
|
| 379 |
+
remove_active_stream_flag_file(session_hash_code)
|
| 380 |
+
remove_active_task_flag_file(session_hash_code)
|
| 381 |
+
|
| 382 |
+
def start_streaming(session_hash_code: str):
|
| 383 |
+
"""Trigger the start flag for active streaming."""
|
| 384 |
+
logging.info(f"[{session_hash_code}] Start button clicked — sending start signal.")
|
| 385 |
+
active_stream_flag = get_active_stream_flag_file(session_hash_code)
|
| 386 |
+
with open(active_stream_flag, "w") as f:
|
| 387 |
+
f.write("1")
|
| 388 |
+
|
| 389 |
+
def _is_stop_requested(session_hash_code) -> bool:
|
| 390 |
+
"""Check if the stop signal was requested."""
|
| 391 |
+
return not os.path.exists(get_active_stream_flag_file(session_hash_code))
|
| 392 |
+
|
| 393 |
|
| 394 |
|
| 395 |
def raise_function():
|