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

Merge pull request #98 from JSchmie/bug_fix_before_v0.2

Browse Source 
Bug fix before v0.2
This commit is contained in:
Jacob Schmieder
2024-05-31 14:40:26 +02:00
committed by GitHub
parent 72f57d5a7f 5ec66effc2
commit 74a37f0732
1 changed files with 45 additions and 85 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scraibe/cli.py
+9 -49
View File
@@ -5,17 +5,12 @@ The function includes arguments for specifying the audio files, model paths,
output formats, and other options necessary for transcription.
"""
import os
from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter
import json
from .autotranscript import Scraibe
from .misc import ParseKwargs
from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter
from whisper.tokenizer import LANGUAGES, TO_LANGUAGE_CODE
from torch.cuda import is_available
from torch import set_num_threads
from .autotranscript import Scraibe
def cli():
"""
@@ -37,21 +32,12 @@ def cli():
parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
group = parser.add_mutually_exclusive_group()
parser.add_argument("-f", "--audio-files", nargs="+", type=str, default=None,
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-type", type=str, default="whisper",
choices=["whisper", "whisperx"],
help="Type of Whisper model to use ('whisper' or 'whisperx').")
parser.add_argument("--whisper-model-name", default="medium",
help="Name of the Whisper model to use.")
@@ -83,7 +69,7 @@ def cli():
parser.add_argument("--verbose-output", type=str2bool, default=True,
help="Enable or disable progress and debug messages.")
parser.add_argument("--task", type=str, default='autotranscribe', # unifinished code
parser.add_argument("--task", type=str, default='autotranscribe',
choices=["autotranscribe", "diarization",
"autotranscribe+translate", "translate", 'transcribe'],
help="Choose to perform transcription, diarization, or translation. \
@@ -104,22 +90,20 @@ def cli():
out_format = arg_dict.pop("output_format")
# seup server arg:
start_server = arg_dict.pop("start_server")
task = arg_dict.pop("task")
if args.num_threads > 0:
set_num_threads(arg_dict.pop("num_threads"))
class_kwargs = {'whisper_model': arg_dict.pop("whisper_model_name"),
'whisper_type':arg_dict.pop("whisper_type"),
'dia_model': arg_dict.pop("diarization_directory"),
'use_auth_token': arg_dict.pop("hf_token")}
'use_auth_token': arg_dict.pop("hf_token"),
}
if arg_dict["whisper_model_directory"]:
class_kwargs["download_root"] = arg_dict.pop("whisper_model_directory")
if not start_server:
model = Scraibe(**class_kwargs)
@@ -166,29 +150,5 @@ def cli():
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__":
cli()