admin/scribe
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

removed gradio related stuff

Browse Source
This commit is contained in:
Schmieder, Jacob
2024-05-30 14:18:33 +00:00
parent 7a388a129a
commit c8b9a7aa93
1 changed files with 35 additions and 75 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scraibe/cli.py
+35 -75
View File
@@ -37,22 +37,9 @@ def cli():
parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter) parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
group = parser.add_mutually_exclusive_group()
parser.add_argument("-f", "--audio-files", nargs="+", type=str, default=None, parser.add_argument("-f", "--audio-files", nargs="+", type=str, default=None,
help="List of audio files to transcribe.") help="List of audio files to transcribe.")
group.add_argument('--start-server', action='store_true',
help='Start the Gradio app.'
'If set, all other arguments are ignored'
'besides --server-config or --server-kwargs.')
parser.add_argument("--server-config", type=str, default=None,
help="Path to the configy.yml file.")
parser.add_argument('--server-kwargs', nargs='*', action=ParseKwargs, default={},
help='Keyword arguments for the Gradio app.')
parser.add_argument("--whisper-model-name", default="medium", parser.add_argument("--whisper-model-name", default="medium",
help="Name of the Whisper model to use.") help="Name of the Whisper model to use.")
@@ -104,9 +91,6 @@ def cli():
out_format = arg_dict.pop("output_format") out_format = arg_dict.pop("output_format")
# seup server arg:
start_server = arg_dict.pop("start_server")
task = arg_dict.pop("task") task = arg_dict.pop("task")
if args.num_threads > 0: if args.num_threads > 0:
@@ -118,76 +102,52 @@ def cli():
if arg_dict["whisper_model_directory"]: if arg_dict["whisper_model_directory"]:
class_kwargs["download_root"] = arg_dict.pop("whisper_model_directory") class_kwargs["download_root"] = arg_dict.pop("whisper_model_directory")
if not start_server: model = Scraibe(**class_kwargs)
model = Scraibe(**class_kwargs) if arg_dict["audio_files"]:
audio_files = arg_dict.pop("audio_files")
if arg_dict["audio_files"]: if task == "autotranscribe" or task == "autotranscribe+translate":
audio_files = arg_dict.pop("audio_files") for audio in audio_files:
if task == "autotranscribe+translate":
task = "translate"
else:
task = "transcribe"
if task == "autotranscribe" or task == "autotranscribe+translate": out = model.autotranscribe(audio, task=task, language=arg_dict.pop(
for audio in audio_files: "language"), verbose=arg_dict.pop("verbose_output"))
if task == "autotranscribe+translate": basename = audio.split("/")[-1].split(".")[0]
task = "translate" print(f'Saving {basename}.{out_format} to {out_folder}')
else: out.save(os.path.join(
task = "transcribe" out_folder, f"{basename}.{out_format}"))
out = model.autotranscribe(audio, task=task, language=arg_dict.pop( elif task == "diarization":
"language"), verbose=arg_dict.pop("verbose_output")) for audio in audio_files:
basename = audio.split("/")[-1].split(".")[0] if arg_dict.pop("verbose_output"):
print(f'Saving {basename}.{out_format} to {out_folder}') print("Verbose not implemented for diarization.")
out.save(os.path.join(
out_folder, f"{basename}.{out_format}"))
elif task == "diarization": out = model.diarization(audio)
for audio in audio_files: basename = audio.split("/")[-1].split(".")[0]
if arg_dict.pop("verbose_output"): path = os.path.join(out_folder, f"{basename}.{out_format}")
print("Verbose not implemented for diarization.")
out = model.diarization(audio) print(f'Saving {basename}.{out_format} to {out_folder}')
basename = audio.split("/")[-1].split(".")[0]
path = os.path.join(out_folder, f"{basename}.{out_format}")
print(f'Saving {basename}.{out_format} to {out_folder}') with open(path, "w") as f:
json.dump(json.dumps(out, indent=1), f)
with open(path, "w") as f: elif task == "transcribe" or task == "translate":
json.dump(json.dumps(out, indent=1), f)
elif task == "transcribe" or task == "translate": for audio in audio_files:
for audio in audio_files: out = model.transcribe(audio, task=task,
language=arg_dict.pop("language"),
out = model.transcribe(audio, task=task, verbose=arg_dict.pop("verbose_output"))
language=arg_dict.pop("language"), basename = audio.split("/")[-1].split(".")[0]
verbose=arg_dict.pop("verbose_output")) path = os.path.join(out_folder, f"{basename}.{out_format}")
basename = audio.split("/")[-1].split(".")[0] with open(path, "w") as f:
path = os.path.join(out_folder, f"{basename}.{out_format}") f.write(out)
with open(path, "w") as f:
f.write(out)
else: # unfinished code
raise NotImplementedError("Currently not Working")
import subprocess
import sys
execute_path = os.path.join(
os.path.dirname(__file__), "app/app_starter.py")
config = arg_dict.pop("server_config")
server_kwargs = arg_dict.pop("server_kwargs")
if not config:
subprocess.run([sys.executable, execute_path,
f"--server-kwargs={server_kwargs}"])
elif not server_kwargs:
subprocess.run([sys.executable, execute_path,
f"--server-config={config}"])
elif not config and not server_kwargs:
subprocess.run([sys.executable, execute_path])
else:
subprocess.run([sys.executable, execute_path,
f"--server-config={config}", f"--server-kwargs={server_kwargs}"])
if __name__ == "__main__": if __name__ == "__main__":