Update modules/whisper/whisper_base.py
Browse files
modules/whisper/whisper_base.py
CHANGED
@@ -229,11 +229,13 @@ class WhisperBase(ABC):
|
|
229 |
mel = whisper.log_mel_spectrogram(whisper.pad_or_trim(whisper.load_audio(file))).to(model.device)
|
230 |
_, probs = model.detect_language(mel)
|
231 |
file_language = ""
|
|
|
232 |
for key,value in whisper.tokenizer.LANGUAGES.items():
|
233 |
if key == str(max(probs, key=probs.get)):
|
234 |
file_language = value.capitalize()
|
|
|
235 |
break
|
236 |
-
|
237 |
transcribed_segments, time_for_task = self.run(
|
238 |
file,
|
239 |
progress,
|
@@ -249,7 +251,7 @@ class WhisperBase(ABC):
|
|
249 |
file_format=file_format,
|
250 |
output_dir=self.output_dir
|
251 |
)
|
252 |
-
files_info[file_name] = {"subtitle": subtitle, "time_for_task": time_for_task, "path": file_path, "lang": file_language, "lang_prob":
|
253 |
|
254 |
## Add output file as txt
|
255 |
file_name, file_ext = os.path.splitext(os.path.basename(file))
|
|
|
229 |
mel = whisper.log_mel_spectrogram(whisper.pad_or_trim(whisper.load_audio(file))).to(model.device)
|
230 |
_, probs = model.detect_language(mel)
|
231 |
file_language = ""
|
232 |
+
file_lang_probs = ""
|
233 |
for key,value in whisper.tokenizer.LANGUAGES.items():
|
234 |
if key == str(max(probs, key=probs.get)):
|
235 |
file_language = value.capitalize()
|
236 |
+
file_lang_probs = str(max(probs, key=probs.get))
|
237 |
break
|
238 |
+
|
239 |
transcribed_segments, time_for_task = self.run(
|
240 |
file,
|
241 |
progress,
|
|
|
251 |
file_format=file_format,
|
252 |
output_dir=self.output_dir
|
253 |
)
|
254 |
+
files_info[file_name] = {"subtitle": subtitle, "time_for_task": time_for_task, "path": file_path, "lang": file_language, "lang_prob": file_lang_probs, "input_source_file": (file_name+file_ext)}
|
255 |
|
256 |
## Add output file as txt
|
257 |
file_name, file_ext = os.path.splitext(os.path.basename(file))
|