azalea/zshrc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[F] Use large-v3 instead of distil-large-v3 for langauge support

Browse Source
This commit is contained in:
Azalea Gui
2024-12-11 01:33:01 -05:00
parent 711962171d
commit fa39a9c45e
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scripts/bin/transcribe
+2 -2
View File
@@ -9,7 +9,7 @@ from faster_whisper import WhisperModel, BatchedInferencePipeline
# model_name = 'deepdml/faster-whisper-large-v3-turbo-ct2' # model_name = 'deepdml/faster-whisper-large-v3-turbo-ct2'
model_name = 'distil-large-v3' model_name = 'large-v3'
# model_name = 'medium.en' # model_name = 'medium.en'
m = WhisperModel(model_name, device="cuda", compute_type="float16") m = WhisperModel(model_name, device="cuda", compute_type="float16")
model = BatchedInferencePipeline(model=m) model = BatchedInferencePipeline(model=m)
@@ -54,7 +54,7 @@ def main():
parser = argparse.ArgumentParser(description="Transcribe audio from a video file and generate an SRT file.") parser = argparse.ArgumentParser(description="Transcribe audio from a video file and generate an SRT file.")
# parser.add_argument("input_file", help="Path to the video file for transcription") # parser.add_argument("input_file", help="Path to the video file for transcription")
parser.add_argument("input_file", nargs="+", help="Path to the video file for transcription") parser.add_argument("input_file", nargs="+", help="Path to the video file for transcription")
parser.add_argument("-l", "--lang", default="en", help="Language code for transcription") parser.add_argument("-l", "--lang", default=None, help="Language code for transcription (e.g. 'en')")
args = parser.parse_args() args = parser.parse_args()
for file in args.input_file: for file in args.input_file:
transcribe(Path(file), args.lang) transcribe(Path(file), args.lang)