run

#!/usr/bin/env python3

import argparse
import datetime
import logging
import os
import sys

from transcribe import whisper

parser = argparse.ArgumentParser(
    prog="run", description="Run transcription generation for sample data"
)

parser.add_argument("--output-dir", help="Path to a directory to write results")
parser.add_argument("--manifest", default="uhec-data.csv", help="Path to data manifest CSV")
parser.add_argument("--threads", help="'--threads' parameter to pass to whisper.cpp")
parser.add_argument("--verbose", help="Print transcript lines as they are generated", action="store_true")
parser.add_argument(
    "--only",
    choices=["whisper", "preprocessing"],
    help="Only run one transcription type",
)

args = parser.parse_args()

# determine where to write results
output_dir = args.output_dir
if output_dir is None:
    output_dir = datetime.date.today().strftime("output-%Y-%m-%d")
if not os.path.isdir(output_dir):
    os.makedirs(output_dir)

# ensure manifest CSV exists
if not os.path.isfile(args.manifest):
    sys.exit(f"manifest file {args.manifest} doesn't exist")

logging.basicConfig(
    filename=os.path.join(output_dir, "transcribe.log"),
    filemode="a",
    format="%(asctime)s,%(msecs)d %(name)s %(levelname)s %(message)s",
    datefmt="%H:%M:%S",
    level=logging.INFO,
)

whisper.run(output_dir, args.manifest, args.threads, args.verbose)
print()
whisper.run_preprocessing(output_dir, args.manifest)