正式whisper

app.py

@@ -331,7 +331,7 @@ def open_folder(folder_path: str):
 
 
 parser = argparse.ArgumentParser()
-parser.add_argument('--whisper_type', type=str, default=WhisperImpl.FASTER_WHISPER.value,
+parser.add_argument('--whisper_type', type=str, default=WhisperImpl.WHISPER.value,
                     choices=[item.value for item in WhisperImpl],
                     help='A type of the whisper implementation (Github repo name)')
 parser.add_argument('--share', type=str2bool, default=False, nargs='?', const=True, help='Gradio share value')

backend/common/audio.py

@@ -1,7 +1,7 @@
 from io import BytesIO
 import numpy as np
 import httpx
-import faster_whisper
+from whisper import audio as whisper_audio
 from pydantic import BaseModel
 from fastapi import (
     HTTPException,
@@ -31,6 +31,6 @@ async def read_audio(
             raise HTTPException(status_code=422, detail="Could not download the file")
         file_content = file_response.content
     file_bytes = BytesIO(file_content)
-    audio = faster_whisper.audio.decode_audio(file_bytes)
+    audio = whisper_audio.decode_audio(file_bytes)
     duration = len(audio) / 16000
     return audio, AudioInfo(duration=duration)

backend/routers/transcription/router.py

@@ -12,7 +12,7 @@
 from datetime import datetime
 from modules.whisper.data_classes import *
 from modules.utils.paths import BACKEND_CACHE_DIR
-from modules.whisper.faster_whisper_inference import FasterWhisperInference
+from modules.whisper.whisper_Inference import WhisperInference
 from backend.common.audio import read_audio
 from backend.common.models import QueueResponse
 from backend.common.config_loader import load_server_config
@@ -41,9 +41,9 @@ def progress_callback(progress_value: float):
 
 
 @functools.lru_cache
-def get_pipeline() -> 'FasterWhisperInference':
+def get_pipeline() -> 'WhisperInference':
     config = load_server_config()["whisper"]
-    inferencer = FasterWhisperInference(
+    inferencer = WhisperInference(
         output_dir=BACKEND_CACHE_DIR
     )
     inferencer.update_model(

modules/utils/audio_manager.py

@@ -2,7 +2,7 @@
 import soundfile as sf
 import os
 import numpy as np
-from faster_whisper.audio import decode_audio
+from whisper.audio import decode_audio
 
 from modules.utils.files_manager import is_video
 from modules.utils.logger import get_logger

modules/utils/logger.py

@@ -14,9 +14,12 @@ def get_logger(name: Optional[str] = None):
             "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
         )
 
-        handler = logging.StreamHandler()
-        # handler.setFormatter(formatter)
+        stream_handler = logging.StreamHandler()
+        stream_handler.setFormatter(formatter)
+        logger.addHandler(stream_handler)
 
-        logger.addHandler(handler)
+        file_handler = logging.FileHandler("webui.log")
+        file_handler.setFormatter(formatter)
+        logger.addHandler(file_handler)
 
     return logger

modules/whisper/base_transcription_pipeline.py

@@ -117,6 +117,8 @@ def run(self,
         """
         start_time = time.time()
 
+        logger.info(f"Run started. audio={audio}")
+
         if not validate_audio(audio):
             return [Segment()], 0
 
@@ -125,6 +127,7 @@ def run(self,
         bgm_params, vad_params, whisper_params, diarization_params = params.bgm_separation, params.vad, params.whisper, params.diarization
 
         if bgm_params.is_separate_bgm:
+            logger.info("Starting background music separation")
             music, audio, _ = self.music_separator.separate(
                 audio=audio,
                 model_name=bgm_params.uvr_model_size,
@@ -145,10 +148,12 @@ def run(self,
             if bgm_params.enable_offload:
                 self.music_separator.offload()
             elapsed_time_bgm_sep = time.time() - start_time
+            logger.info("BGM separation completed")
 
         origin_audio = deepcopy(audio)
 
         if vad_params.vad_filter:
+            logger.info("Applying VAD")
             progress(0, desc="Filtering silent parts from audio..")
             vad_options = VadOptions(
                 threshold=vad_params.threshold,
@@ -168,7 +173,9 @@ def run(self,
                 audio = vad_processed
             else:
                 vad_params.vad_filter = False
+            logger.info("VAD completed")
 
+        logger.info("Starting transcription")
         result, elapsed_time_transcription = self.transcribe(
             audio,
             progress,
@@ -177,6 +184,7 @@ def run(self,
         )
         if whisper_params.enable_offload:
             self.offload()
+        logger.info("Transcription completed")
 
         if vad_params.vad_filter:
             restored_result = self.vad.restore_speech_timestamps(
@@ -189,6 +197,7 @@ def run(self,
                 logger.info("VAD detected no speech segments in the audio.")
 
         if diarization_params.is_diarize:
+            logger.info("Running diarization")
             progress(0.99, desc="Diarizing speakers..")
             result, elapsed_time_diarization = self.diarizer.run(
                 audio=origin_audio,
@@ -198,6 +207,7 @@ def run(self,
             )
             if diarization_params.enable_offload:
                 self.diarizer.offload()
+            logger.info("Diarization completed")
 
         self.cache_parameters(
             params=params,
@@ -261,6 +271,8 @@ def transcribe_file(self,
                 "highlight_words": True if params.whisper.word_timestamps else False
             }
 
+            logger.info(f"Transcribing files: {files} input_folder={input_folder_path}")
+
             if input_folder_path:
                 files = get_media_files(input_folder_path, include_sub_directory=include_subdirectory)
             if isinstance(files, str):
@@ -270,6 +282,7 @@ def transcribe_file(self,
 
             files_info = {}
             for file in files:
+                logger.info(f"Processing file {file}")
                 transcribed_segments, time_for_task = self.run(
                     file,
                     progress,
@@ -312,6 +325,7 @@ def transcribe_file(self,
             result_str = f"Done in {self.format_time(total_time)}! Subtitle is in the outputs folder.\n\n{total_result}"
             result_file_path = [info['path'] for info in files_info.values()]
 
+            logger.info("File transcription completed")
             return result_str, result_file_path
 
         except Exception as e:

modules/whisper/data_classes.py

@@ -368,7 +368,7 @@ def to_gradio_inputs(cls,
                          available_langs: Optional[List] = None,
                          available_compute_types: Optional[List] = None,
                          compute_type: Optional[str] = None):
-        whisper_type = WhisperImpl.FASTER_WHISPER.value if whisper_type is None else whisper_type.strip().lower()
+        whisper_type = WhisperImpl.WHISPER.value if whisper_type is None else whisper_type.strip().lower()
 
         inputs = []
         if not only_advanced:

modules/whisper/faster_whisper_inference.py

@@ -64,12 +64,14 @@ def transcribe(self,
         elapsed_time: float
             elapsed time for transcription
         """
+        logger.info("Transcribe called")
         start_time = time.time()
 
         params = WhisperParams.from_list(list(whisper_params))
 
         if params.model_size != self.current_model_size or self.model is None or self.current_compute_type != params.compute_type:
             self.update_model(params.model_size, params.compute_type, progress)
+            logger.info(f"Model set to {params.model_size} ({params.compute_type})")
 
         segments, info = self.model.transcribe(
             audio=audio,
@@ -112,6 +114,7 @@ def transcribe(self,
             segments_result.append(Segment.from_faster_whisper(segment))
 
         elapsed_time = time.time() - start_time
+        logger.info(f"Transcription done in {elapsed_time:.2f}s")
         return segments_result, elapsed_time
 
     def update_model(self,
@@ -133,6 +136,7 @@ def update_model(self,
         progress: gr.Progress
             Indicator to show progress directly in gradio.
         """
+        logger.info(f"Loading model {model_size} with compute_type {compute_type}")
         progress(0, desc="Initializing Model..")
 
         model_size_dirname = model_size.replace("/", "--") if "/" in model_size else model_size
@@ -163,6 +167,7 @@ def update_model(self,
             compute_type=self.current_compute_type,
             local_files_only=local_files_only
         )
+        logger.info("Model initialized")
 
     def get_model_paths(self):
         """

modules/whisper/whisper_factory.py

@@ -91,8 +91,8 @@ def create_whisper_inference(
                 uvr_model_dir=uvr_model_dir
             )
         else:
-            return FasterWhisperInference(
-                model_dir=faster_whisper_model_dir,
+            return WhisperInference(
+                model_dir=whisper_model_dir,
                 output_dir=output_dir,
                 diarization_model_dir=diarization_model_dir,
                 uvr_model_dir=uvr_model_dir

tests/test_bgm_separation.py

@@ -18,7 +18,6 @@
     "whisper_type,vad_filter,bgm_separation,diarization",
     [
         (WhisperImpl.WHISPER.value, False, True, False),
-        (WhisperImpl.FASTER_WHISPER.value, False, True, False),
         (WhisperImpl.INSANELY_FAST_WHISPER.value, False, True, False)
     ]
 )
@@ -39,7 +38,6 @@ def test_bgm_separation_pipeline(
     "whisper_type,vad_filter,bgm_separation,diarization",
     [
         (WhisperImpl.WHISPER.value, True, True, False),
-        (WhisperImpl.FASTER_WHISPER.value, True, True, False),
         (WhisperImpl.INSANELY_FAST_WHISPER.value, True, True, False)
     ]
 )

tests/test_diarization.py

@@ -17,7 +17,6 @@
     "whisper_type,vad_filter,bgm_separation,diarization",
     [
         (WhisperImpl.WHISPER.value, False, False, True),
-        (WhisperImpl.FASTER_WHISPER.value, False, False, True),
         (WhisperImpl.INSANELY_FAST_WHISPER.value, False, False, True)
     ]
 )

tests/test_transcription.py

@@ -88,7 +88,6 @@ def run_asr_pipeline(
     "whisper_type,vad_filter,bgm_separation,diarization",
     [
         (WhisperImpl.WHISPER.value, False, False, False),
-        (WhisperImpl.FASTER_WHISPER.value, False, False, False),
         (WhisperImpl.INSANELY_FAST_WHISPER.value, False, False, False)
     ]
 )

tests/test_vad.py

@@ -13,7 +13,6 @@
     "whisper_type,vad_filter,bgm_separation,diarization",
     [
         (WhisperImpl.WHISPER.value, True, False, False),
-        (WhisperImpl.FASTER_WHISPER.value, True, False, False),
         (WhisperImpl.INSANELY_FAST_WHISPER.value, True, False, False)
     ]
 )